inte 001-028 ch01 - World Banksiteresources.worldbank.org/INTRANETTRADE/Resources/C13.pdf · ubp_PTAPD_275-306.qxd:inte_001-028_ch01 7/5/11 12:28 PM Page 275. The Concept of Labor

275

13

Of all the liberalization dimensions of internationaltrade, the opening of labor markets is probably the mostsensitive. As a result, agreements to open (generally, onlypartially) labor markets are not as plentiful as other liber-alization agreements, and they are typically more restric-tive. They are also less well surveyed. This chapter presentsan overview of provisions for opening labor markets thatare found in preferential trade agreements (PTAs).

Movement of labor is one of the four fundamental eco-nomic freedoms, along with free movement of goods,services, and capital. Of the four, it has met with the leastreceptivity on the part of countries in the internationaleconomy, whether developed or developing. Even themost spirited free traders—Chile; Singapore; the UnitedKingdom; and Hong Kong SAR, China, for example—have been reticent about opening their borders moreto admit labor from abroad. These economies and manyothers shy away from significant opening for natural per-sons from other countries, even in the face of labor short-ages at home.

The contrast between the desire to promote capitalmobility and investment flows and the reluctance to envis-age corresponding labor mobility is stark. More than 2,800bilateral investment treaties have been signed to date, butnothing equivalent exists in the area of labor (Vis-Dunbarand Nikiema 2009). The number of trade agreements cov-ering services is growing rapidly, yet willingness to incor-porate meaningful provisions on labor mobility into theservices package is limited, and most agreements containvery modest market access opportunities for foreignworkers. Several recent free trade agreements (FTAs) con-tain no provisions at all in this area. Leaving aside formalbilateral and regional trade agreements, it is extremely dif-ficult to determine the number of bilateral agreementsworldwide that incorporate arrangements for temporaryworker programs. There appears, however, to be notmore than a handful. Information is scarce, and no single

institution has been designated as a repository of agree-ments and promoter of labor mobility. And whereas mostgovernments sponsor investment promotion agencies toencourage inward flows of capital and foreign directinvestment, there is no similar institution for workers. Innearly all countries, the agency that deals with the influx offoreign labor is the immigration authority, whose concernis to regulate and restrict, not to promote. To complicatematters further, immigration authorities are primarilyfocused on setting rules for permanent rather than tempo-rary migration. Temporary migration, however, is theobject of international trade policy and is the focus of thisdiscussion.

The problems with the present situation are twofold.The first has to do with its unbalanced nature. Developedeconomies have a comparative advantage in the export ofcapital and thus benefit tremendously from the opennessof capital markets and the welcoming character of mostinvestment regimes. That kind of receptiveness does notexist for labor movements. Developing countries have acomparative advantage in the export of their people, butthey are constrained from realizing the gains from tradethat they might otherwise enjoy.

The second problem is that the entire world suffersfrom a loss of potential income that could be realizedthrough greater mobility. Depending on what assump-tions are made by researchers, the potential gains could bequite substantial and could easily surpass the combinedgains anticipated from freer trade in agriculture and man-ufactured goods—as currently proposed in the DohaDevelopment Round sponsored by the World TradeOrganization (WTO). Nevertheless, proposals for greatermarket access for foreign workers are limited, and this hasbeen a central obstacle to progress in the services compo-nent of Doha Round negotiations.

It has been argued by many, including, prominently,Lant Pritchett and L. Alan Winters, that greater mobility of

LABOR MOBILITY

Sherry Stephenson and Gary Hufbauer

ubp_PTAPD_275-306.qxd:inte_001-028_ch01 7/5/11 12:28 PM Page 275

The Concept of Labor Mobility

In international services trade, labor mobility is conceptu-alized as the temporary movement of natural persons andis categorized as mode 4. Article I.2 (d) of the WTO Gen-eral Agreement on Trade in Services (GATS) defines mode4 as the supply of a service “by a service supplier of oneMember, through presence of natural persons of a Memberin the territory of any other Member.” (See box 13.1 forestimates of the size of this mode.) A natural person ofanother member is defined as

a natural person who resides in the territory of that other

Member or any other Member, and who under the law of

that other Member:

(i) is a national of that other Member; or

(ii) has the right of permanent residence in that other

Member . . .” (Article XVIII[k])

Temporary versus Permanent Workers

For services trade, and for our purposes here, labor mobil-ity is understood as the movement of workers to carry outemployment in another country for a time-limited period.Although the term “temporary” is not defined underGATS, the notion of moving in order to work for a limitedperiod, as opposed to moving with the intention of emi-grating permanently, is what distinguishes mode 4. This isaffirmed in the GATS “Annex on Movement of NaturalPersons Supplying Services under the Agreement,” whichspecifies that GATS shall not apply “to measures affectingnatural persons seeking access to the employment marketof a Member, nor shall it apply to measures regarding citi-zenship, residence or employment on a permanent basis.”All subsequent trade agreements, following the WTOapproach, consider only the temporary movement ofworkers, but governments have been unwilling to define inprecise terms what period of time is meant by “temporary.”The official statistical definition of temporary migration isa stay of less than one year, but for trade policy purposes, atemporary stay can vary anywhere from a few weeks to afew years, depending on the commitments governmentsare prepared to undertake. This lack of precision has beenboth a strength and a weakness in defining mode 4 treat-ment within trade agreements.

The great political sensitivity surrounding interna-tional labor mobility is not helped by the very frequentconfusion, in both statistical analysis and politicaldebate, between temporary and permanent migration.Immigration authorities deal with both simultaneously,and at times the character of “temporary” labor movementis disregarded and all immigrants are treated as though

labor would be the first-best development promotion strat-egy (Pritchett 2006; Winters 2008). Pritchett writes that it ishard to imagine a policy more directly at odds with povertyreduction or pro-poor growth objectives than one limitingthe demand for lower-skilled labor. This limitation can beviewed as the principal way that rich countries are cur-rently inhibiting the development possibilities of poorercountries—much more than through restrictive agricul-tural policies or nontariff barriers.

In such a challenging and often hostile environmentfor labor mobility, what options might developing coun-tries have for increasing the scope for the movement oftheir workers? Given the impasse in the Doha Round andthe lack of any progress on services in multilateral nego-tiations for the past several years, preferential tradeagreements might offer a more promising channel forgreater labor mobility, even if among a limited numberof partners. Other options that have been relativelyunexplored to date may also be available, such as thepromotion of circular migration through temporaryworker agreements—time-bound instruments that allowgreater flexibility for both labor-sending and labor-receiving countries.

In this chapter, we do not delve into the Doha quagmirebut, rather, explore these other options. We first discuss theconcept and magnitude of labor mobility and the potentialbenefits from greater liberalization. We then review thevarious ways in which members of PTAs have treated theissue of labor mobility to assess whether these preferentialagreements have effectively promoted temporary entry.Only PTAs between developed and developing countriesare examined.

The questions we attempt to answer are the following:Which developed countries are more amenable to greateropenness for natural persons, and what are the possiblereasons? Can recent PTAs be emulated? What have tem-porary worker programs in bilateral and plurilateralagreements with developing-country partners achieved,and could such agreements usefully supplement thePTA approach? After exploring these issues, we look atpolicy suggestions that might be implemented by theWorld Bank to promote labor mobility for its developing-country members.

We recognize that developed countries are currentlyexperiencing very high levels of unemployment and thatthese conditions will probably continue through 2011and possibly into 2012. Accordingly, political resistanceto all forms of labor mobility is extremely high. Thischapter, however, is written with a view to the longerterm, when normal economic conditions will have beenrestored.

276 Sherry Stephenson and Gary Hufbauer


they were seeking permanent status. Moreover, inside thehost country, the line between permanently and temporar-ily resident migrants often becomes blurred.

Categories of Labor Included in Trade Agreements

Although the GATS text does not define specific categoriesof labor, WTO members have accepted four widely usedcategories for the purpose of inscribing commitmentsunder mode 4. These categories are not comprehensive, asthey cover only skilled professionals. In a few recent tradeagreements, as we will see in the next section, countrieshave begun to move beyond this limited range of cate-gories to broaden their consideration of labor categoriesfor market access. The four traditional mode 4 categoriesare the following:

• Business visitors and salespersons (BVs). Foreign nation-als who travel abroad to negotiate the sale of a serviceor to explore the possibility of making a foreign directinvestment (of establishing a commercial presence,in GATS terminology) for their company in the desti-nation country. Their main purpose is to facilitatefuture transactions rather than actually to carry outtransactions.

• Intracorporate transferees (ICTs). Employees of a foreignservices company that has set up a commercial presenceabroad and that transfers these employees to its foreignlocation.

• Independent professionals (IPs). Self-employed personswho are supplying a service to a company or an individ-

ual in a host country. In most trade agreements, thesehave been limited to professional workers, but commit-ments can also be extended to lower-skilled categoriesof workers.

• Contractual services suppliers (CSSs). Employees of aforeign services company with no local presence orcommercial presence in the host country who areengaged under contract to provide a service to a firm inthe destination country.1

Developing countries’ interest in promoting greaterlabor mobility most concerns the independent profes-sional and contractual services supplier categories,rather than employees of multinational corporations(MNCs). This is because most developing countries,with notable exceptions such as Brazil, China, and India,have not yet become home bases for MNCs. Greater flex-ibility in the independent professional and contractualservices supplier categories would allow most developingcountries to send a larger number of professionalsabroad for temporary employment. The business visitorand intracorporate transferee categories are of interestto successful emerging countries such as Brazil, China,and India.

Potential Economic Gains from Greater Labor Mobility

The fact that we cannot accurately know the real statisti-cal importance of temporary workers in the world econ-omy is secondary to the fact that impediments to labor

Labor Mobility 277

Box 13.1. Labor Mobility in Statistical Terms

Historically, one of the main ways that temporary labor migration has been captured in the data is through recorded “transfers andpayments” in balance of payments statistics; this category is what we and others term “remittances.” According to the InternationalMonetary Fund (IMF) Balance of Payments Manual, “remittances” mainly comprise “compensation of employees” and “personaltransfers” (IMF 2010). Transactions are recorded in the balance of payments when money is paid by residents to nonresidents, orvice versa. Determining the magnitude of mode 4 (temporary movement of services providers) by examining transactions isproblematic, since remittances are made by both temporary and permanent migrants and by workers in the manufacturing andagricultural segments of the economy, as well as in services sectors. In addition, payments to undocumented foreign workers arenot captured statistically when they are spent in the country in which the person works (see also Carzaniga 2008).

Looking at volume rather than value is hardly more satisfactory. The Organisation for Economic Co-operation and Development(OECD) collects information on the number of temporary foreign workers in surveys of firms, visas, border crossings, and so on.These data cover intracorporate transferees and other temporary workers. They are, however, far from a perfect match for mode 4.Furthermore, the scope may differ from one country to another, and in any event, figures are only available for a subgroup of OECDmembers.

Karsenty (2000) calculated services trade through mode 4 to be no more than 1 to 2 percent of total two-way trade in services.Applying this range to the most recent trade data available (2008) would put the value of services trade through mode 4somewhere between US$70 billion and US$150 billion annually. Those figures may underestimate the actual value of this form ofservices trade, since remittances alone in 2007 amounted to more than US$200 billion. Remittances, however, likely overestimatemode 4 services trade.

Source: Karsenty 2000; OECD data.


for native workers is shown by area ACDE. The gain forcapitalists is shown by area EABD, with most of this gaincoming from the loss for native workers. Since the gain forcapitalists is larger than the loss for native workers, the lib-eralization of mode 4 leads to an overall gain, shown byarea ABC.

Effect on developing countries. The effect of the liberal-ization of mode 4 on developing countries is the exactopposite to that for developed countries. With restrictionson mode 4, the equilibrium in the labor market is at pointB in figure 13.3. After liberalization, the equilibrium pointmoves to point A, reflecting an increase in the wage perhour and a decrease in the number of hours worked.

As will be apparent later, the gains for migrants indeveloped countries are much larger than the loss thattheir departure inflicts on developing countries. Nonmi-grant workers also experience gains, shown by area ACDEin figure 13.3, since the wage rate has increased in devel-oping countries. But nonmigrant capitalists experience avery large loss, shown by area ABDE (most of the loss cor-responds to the wage gain for nonmigrant workers).Because the loss for nonmigrant capitalists is larger thanthe gain for nonmigrant workers, the group of nonmi-grants as a whole experiences an overall loss of income,shown by area ABC. In other words, the effect on totalwelfare of liberalizing mode 4 is negative for nonmigrantsin developing countries. Income per capita, however, islikely (although not guaranteed) to rise as marginal pro-ductivity increases.

Overall outcome. Migrants lose their erstwhile wages indeveloping countries but enjoy larger wages in developedcountries. They therefore experience a gain, measured bythe wage difference between the destination and sourcecountries.

mobility suppress trade to the disadvantage of everyone,but particularly developing countries. Goods move freelyin response to price differentials, and capital flows effort-lessly around the globe in response to profit and interestrate differentials, but workers are not allowed to movereadily in response to wage differentials. Consequently,very large wage differentials exist in the world today, asshown in figure 13.1. The benefit to be derived from theexploitation of comparative advantage is directly propor-tional to the size of wage, price, or profit differences priorto trade or investment liberalization; thus, considerablegains could be realized if workers were permitted toexploit wage differentials among countries. (See box 13.2for estimates of these gains.)

Theoretical Model of the Distributional Effects ofMode 4 Liberalization

Like trade in goods, labor mobility can create losers aswell as winners. In the overall balance, gains usuallyexceed losses by a wide margin, but political sensitivitiesfocus on those who lose. In simple theoretical terms,migration can be modeled as an increase of supply in thelabor markets of developed countries and a decrease ofsupply in developing countries. Here, we use that frame-work to examine the effects of those supply changes onthe incomes of capitalists and workers, in both the send-ing and the host countries, and on the incomes of themigrants themselves.

Effect in developed countries. Given the restrictions onlabor mobility, the equilibrium in the labor market is atpoint A in figure 13.2. After liberalization, the equilibriummoves to point B, reflecting an increase in the number ofhours worked and a decrease in the wage per hour. The loss


Figure 13.1. Theoretical Gains from Liberalization of Mode 4

wagedifferential

restrictions onlabor mobility

potential gain

productivity/wages indevelopedcountries

productivity/wages indevelopingcountries

number of workers indeveloped countries

number of workers indeveloping countriesnumber of

migrants


According to the theoretical model, the liberalization ofmode 4 has the following distributional consequences:

• In developed countries, most of the gains for capitalistsare balanced by losses to native workers.

• In developing countries, most of the losses to capitalistsare mirrored by gains to nonmigrant workers.

• In developed countries, the gains for capitalists arelarger than the losses for native workers. Therefore, totalincome in developed countries rises.

Labor Mobility 279

Box 13.2. Quantitative Estimates of Overall Gains from Greater Labor Mobility

Complete liberalization of mode 4 would result in very large gains.Hamilton and Whalley (1984) use a partial equilibrium (PE) model and 1977 data to estimate the benefits from the completeelimination of all immigration restrictions, for skilled and unskilled labor alike. The potential gains are enormous, ranging from 60to almost 205 percent of world gross domestic product (GDP). Millions of workers would move from low-productivity to high-productivity jobs in countries with high salaries, until wages in labor-sending and labor-receiving countries equalized. Iregui (1999)revisits the question using a computable general equilibrium (CGE) model and more precise measures of elasticities and populationcharacteristics. Here again, the gains are large, ranging from 15 to 67 percent of world GDP. Moses and Letnes (2004), using moreprecise values for productivities, confirm large gains, ranging from 4.3 to about 112 percent of world GDP in 1977. According tothese authors, the ‘’most reasonable’’ gain would be 7.5 percent of world GDP.

The large differences between these estimates, both within and between studies, can be explained by the differences inmodeling frameworks (partial versus general equilibrium) and assumed parameters. Some estimates assume that migrants canachieve the average productivity of workers in the destination country; others assume that additional education and training will beneeded.

Gains from less than complete liberalization of mode 4 are still large.Because full liberalization is politically unacceptable, some economists have estimated the potential outcome of more modestliberalization of mode 4. Moses and Letnes (2004) estimate the gains from eliminating 10 percent of the wage inequality betweencountries and find that potential gains would still be large, corresponding to around 2.2 percent of world GDP. Walmsley andWinters (2002) estimate the potential gain from a 3 percent increase in the workforce in developed countries, a movement of 14.2million workers, and a 50 percent increase in the current number of immigrants in developed countries at US$156 billion in 2002,representing 0.6 percent of world GDP. World Bank (2006) reaches a very similar result.

Most of the gains come from the movement of unskilled labor.According to Iregui (1999), the potential gains from the migration of skilled labor only are much smaller: 3 to 11 percent of worldGDP, in comparison with 13 to 59 percent for all skills. Walmsley and Winters (2002) show that the potential gain from themovement of unskilled workers would account for US$110 billion, or 70 percent of the total. This reflects the fact that inequality inwages worldwide is larger for unskilled than for skilled workers.

Source: Annex table 13A.1.

Figure 13.2. Theoretical Effect on DevelopedCountries of Liberalization of Mode 4

number of hours worked

demand

A

BC

D

Eloss for native workers overall

gain for natives

supply(liberalization)

supply(with restrictions)

wages

Figure 13.3. Theoretical Effect on DevelopingCountries of Liberalization of Mode 4

number of hours worked

demand

A

BC

D

Egain for nonmigrant

workersoverall

loss for nonmigrants

supply(liberalization)

supply(with restrictions)

wages


between 1990 and 2008, rising from US$69 billion toUS$397 billion (adjusted for inflation). In 2007, migrantcompensation and remittances accounted for around0.7 percent of world GDP, but for developing countries, the relative importance of remittances in GDP in 2007 wasmuch higher. Remittances were 2.1 percent of the GDP ofdeveloping countries as a whole, but 1.9 percent of theGDP of middle-income countries and 5.8 percent of theGDP of the least-developed countries (a UN category).

An increasing share of remittances goes to developingcountries, which accounted for 46 percent of this flow in1990 but for 76 percent by 2007. It is estimated that remit-tances touch 1 in 10 people worldwide. Dependence onremittances is especially high in certain countries. Themain receiving countries in absolute terms are India(US$27 billion), China (US$26 billion), Mexico (US$25billion), and the Philippines (US$17 billion). For manysmaller countries, remittances represent a very large frac-tion of GDP, accounting for more than 36 percent of theGDPs of Moldova and Tajikistan and about 25 percent ofthe GDPs of Guyana, Honduras, and Lesotho.

Mixed picture in developed countries. Outcomes ofmigration for the developed countries are mixed, althoughslightly positive. Workers, especially unskilled ones, faceincreased competition from migrants and see their wagesdecline. For example, Hatton and Williamson (1998) esti-mate that in 1910, American wages would have been 11 to14 percent higher in the absence of the immigration wavethat set in after 1870. Borjas (1999) calculates that immi-gration to the United States between 1980 and 1998resulted in a decrease in native wages amounting to1.9 percent of GDP and that the losses were concentratedamong low-skilled U.S. workers, whereas skilled workersactually benefited from immigration. Immigrationreduced the wages of native high-school dropouts in theUnited States by 8.9 percent between 1980 and 2000 butincreased the return to capital by 2 percent of GDP. Thenet gain from the 1980–98 migration wave for all U.S.natives is the difference between the decrease in wagesand the increase in returns to capital, or 0.1 percent ofU.S. GDP per year over the period. This net gain repre-sents about US$10 billion a year, accounting for about5 percent of U.S. economic growth over a 20-year period.

Moses and Letnes (2004) find the same pattern in thecase of a 10 percent elimination of wage inequality. Theycalculate that liberalization of this magnitude wouldreduce wages in developed countries by 3.1 percent,while increasing the return to capital by 7.2 percent.Walmsley and Winters (2002) reach similar results in thecase of a 3 percent increase in the workforce of devel-oped countries: that scenario leads to a 0.8 percent

• In developing countries, the losses for capitalists arelarger than the gains for nonmigrant workers. There-fore, total income in developing countries falls.

Distributional Effects of Mode 4 Liberalization

The theoretical and empirical prediction of large gainsfrom full or partial liberalization of mode 4 outlined inbox 13.2 do not hide the fact that labor mobility will havedistributional consequences. Migrants are the main win-ners; the results for natives in both the sending and the hostcountries are mixed.

Gains for migrants. Walmsley and Winters (2002) calcu-late that benefits to migrants (US$171 billion) actuallyaccount for more than the total gain from increased labormobility (US$156 billion). Total gains are smaller than thegains to migrants because of the losses to the sendingcountries, discussed below.

Losses for developing countries, before remittances. Thedeparture of migrants reduces the number of workers inthe sending countries, which increases hourly wages ofnonmigrant workers but diminishes total output. Walms-ley and Winters (2002) calculate that Brazil would see itswelfare reduced by US$7 billion if the workforce going todeveloped countries increased by 3 percent, and Chinawould experience a decline of US$2 billion, notwith-standing the compensation received from remittances.The authors’ calculations suggest that unskilled workersin India would see a wage increase of 0.7 percent and thatskilled workers in Mexico would enjoy an increase of4.5 percent. Returns to capital would, however, decreaseby, for example, 0.4 percent in Mexico. Exploring a moreextreme scenario, Moses and Letnes (2004) arrive at simi-lar results. In their calculations, a 10 percent eliminationof wage inequality leads to an 11.4 percent increase in thewages of nonmigrant workers in the poorest countries in1998, while the return to capital in those countries fallslike a stone, by 21 percent.

The importance of remittances for developing countries.If the gains to migrants themselves are included in theoverall balance sheet for developing countries, the pic-ture changes completely. (Pritchett 2006 makes thispoint.) When the gains to migrants are combined withthe national income losses to the sending countries, thedeveloping countries experience a significant gain inplausible scenarios—the equivalent of 1.8 percent of theirgross domestic product (GDP), according to the WorldBank’s Global Economic Prospects 2006, which exploresthe “3 percent scenario.”

World Bank estimates of global remittances show thatglobally, compensation and remittances increased sixfold



decrease in U.S. and European wages and a 0.8 percentincrease in return to capital in the United States. The WorldBank’s Global Economic Prospects 2006 study shows that inthe 3 percent scenario, the incomes of all natives combinedin developed countries would rise by 0.4 percent (WorldBank 2006).

Labor Mobility in Preferential TradeAgreements

A stalemate on services liberalization at the multilaterallevel has clouded the Doha Round for the past severalyears, with no moves to put new services offers on thetable or to improve existing ones made since the end of2005.2 In contrast, an increasing amount of activity hastaken place at the regional level, with the negotiation ofnumerous PTAs, a large number of which have incorpo-rated mode 4 as part of the package.3

The members of some recent PTAs have acceptedgreater labor mobility at the regional level. Althoughprogress is still relatively modest, interesting initiativeshave been taken. Developing countries in the Americasand in Asia have entered into several free trade agree-ments that contain provisions to facilitate procedures for temporary labor movement and open up marketaccess opportunities. Some agreements include guaran-teed numerical quotas for certain categories of skilledlabor. In this section, we compare approaches to labormobility in PTAs. Only PTAs between developed anddeveloping economies, and only those negotiated since theentry into force in January 1994 of the North AmericanFree Trade Agreement (NAFTA), which signaled an era ofdeeper and more comprehensive preferential trade agree-ments, are considered.

Our discussion is divided along geographic lines, distin-guishing between PTAs negotiated by the United States,Canada, the European Union (EU), Japan, and Australiaand New Zealand. Annex tables 13A.2–13A.6 summarizethe salient provisions of the PTAs that are examined.

Before delving into the details, a broad overview maybe useful. NAFTA and other first-generation U.S. PTAsallowed limited mobility for professional workers, but sec-ond-generation U.S. PTAs are quite restrictive, as a resultof congressional opposition. Canadian PTAs are muchmore liberal for skilled workers, as well as professionals,and contain some innovative provisions. Early EU agree-ments with developing countries did not have provisionsallowing labor mobility because the subject was reservedto the competence of member states, but more recent EUagreements with Chile and the countries of the CaribbeanForum of African, Caribbean, and Pacific (ACP) States

(CARIFORUM) do permit limited access. Japanese PTAsallow the usual professional categories and contain innova-tive provisions for semiskilled workers. Australia and NewZealand have negotiated highly innovative agreementswith China and Chile. By contrast, the Trans-Pacific Strate-gic Economic Partnership Agreement between BruneiDarussalam, Chile, New Zealand, and Singapore is quiterestrictive. This feature may ease future accession by Aus-tralia, Japan, the United States, and Vietnam, but it doesnothing for labor mobility among the Pacific members.

PTAs Negotiated by the United States and Canada

NAFTA was the pioneer agreement and template formany subsequent PTAs. It contains a chapter entitled“Temporary Movement of Business Persons,” designed tofacilitate temporary entry to member countries for business people involved in goods or services trade orin investment activities. The categories defined underNAFTA are traders and investors, business visitors, intra-corporate transferees, and professionals. There is no limiton the number of visas for business visitors, and a workpermit is not required. According to Martin and Lowell(2008), the novel migration component of NAFTA is theTrade NAFTA, or TN, visa (see Stephenson 2007 for asummary). This visa was uncapped in 1994 for Canadiansand has been uncapped for Mexicans since 2004. Whenproof of a job offer is demonstrated, the TN visa permitsemployment for one year, with unlimited renewal.

In addition to the chapter on temporary entry, NAFTA,like subsequent agreements with a similar structure, contains an annex on professionals that is specificallytargeted at professional services suppliers. These annexesare intended to promote the development of mutuallyacceptable standards and criteria for licensing and certi-fication of professional services suppliers, on the basis offactors such as educational background, qualifyingexaminations, and experience. In addition, the NAFTAannex encourages members to submit recommendationsfor furthering the process of mutual recognition. A qual-ifying list of 62 professions is set out in an appendix tothe agreement; applicants must fulfill the necessary qual-ification requirements. The United States originallyplaced a quota of 5,500 per year on the number of pro-fessionals who could be admitted from Mexico, but thatquota has been eliminated.

Besides NAFTA, the United States has negotiated severalother bilateral free trade agreements with developing coun-tries. The agreements selected for examination are thosewith Chile, the five-country Dominican Republic–CentralAmerica Free Trade Agreement (CAFTA–DR), Morocco,

Labor Mobility 281


negotiate greater labor mobility in trade agreements withthe United States.

In Canada, the situation has evolved in the oppositedirection (see annex table 13A.3). Interestingly, and per-haps because of pressures from the private sector andapparent labor shortages in the Canadian market priorto the current economic crisis, the government has negoti-ated recent FTAs that go quite far toward providingincreased access not only for professionals but also for semi-skilled foreign workers. Although the FTA that Canadanegotiated with Chile in 1997 looks very much like NAFTA,with the only categories of workers covered being investors,traders, business visitors, intracorporate transferees, andprofessionals, it is notable in that no numerical limits wereplaced on 72 of these categories of professional labor.

Strikingly, the two very recent FTAs negotiated byCanada with Colombia (2008) and Peru (2009) go muchfarther. They cover all professional categories, with nonumerical limits and no specified length of stay, meaningthat visas could in theory be renewed indefinitely. For thefirst time, they also expand coverage of worker categoriesbeyond highly trained professionals to include “techni-cians.” In both the Colombia and Peru FTAs, Canada haslisted 50 categories of technicians to be admitted into theCanadian market with no specified length of stay. Theseworkers must have a high school degree, with two yearsof technical training. Technician categories include,among others, mechanics, construction inspectors, foodand beverage supervisors, chefs, plumbers, and oil andgas well drillers. This recent development constitutes amajor step forward for the expansion of temporary entryin trade agreements.

PTAs Negotiated by the European Union

In this section, we examine PTAs between the EU and thirdcountries (see annex table 13A.4) Total labor mobility isguaranteed within the EU itself, although only after 10years for some of the newest members.

The form of PTAs negotiated by the EU differs fromthat pioneered by the United States. Provisions for servicesand investment liberalization are set out in a title (or sec-tion) of the EU agreements entitled “Trade in Services andEstablishment.” The European Commission does not as yethave negotiating authority from the EU member states inall service areas.5 The European Commission consequentlyalways follows a positive-list approach in its trade agree-ments, with lists of commitments attached to the main textof the agreements. Thus, in terms of market access, mode 4is brought within the scope of EU PTAs in a way similar tothat followed under GATS. Categories of workers included

Peru, and Singapore (annex table 13A.2). Bilateral agree-ments with Colombia, the Republic of Korea, and Panamahave been finalized but are awaiting ratification by the U.S.Congress.

Under the agreements with Chile and Singapore, bothof which were concluded in 2002 and entered into forcein 2004, labor mobility was expanded slightly for profes-sional workers, and a path to a special visa for profession-als (the H-1B1 visa) was created. The visa provided for aninitial stay of 18 months, but with unlimited extensions.In addition, an annual quota of 1,400 visas for profes-sionals from Chile was granted, as was an annual quota of5,400 visas for professionals from Singapore, on top ofthe fixed total of H-1B visas from all countries. The newvisa category created under these agreements is meant fortemporary migrants, for stays of up to 18 months ini-tially, but with the possibility of unlimited extensions.

In brief, the current provisions governing labor move-ment to the United States under NAFTA and the agree-ments with Chile and Singapore are as follows:

• NAFTA: TN visa; uncapped for both Canadians andMexicans

• Chile FTA: H-1B1 visa; capped at 1,400 professionals• Singapore FTA: H-1B1 visa; capped at 5,400 profes-

sionals.

As mentioned, these visa numbers are additionalto whatever entries occur under other visa categories,most importantly, the H-1B visa for skilled workers andprofessionals.

Unfortunately, the opposition of the U.S. Congress tothese arrangements, and in particular to the agreementswith Chile and Singapore, was loud and clear. Key mem-bers of Congress objected that the trade agreements hadencroached on the realm of immigration matters. As aconsequence of this outcry, no free trade agreement nego-tiated by the United States since 2002 has containeda chapter to facilitate the temporary movement of skilledworkers.4 Thus, the FTAs with Morocco, with theCAFTA–DR members, and with Peru, like those negoti-ated with Colombia, Korea, and Panama, contain nochapter on temporary entry. Each does contain an annexon professionals, with objectives similar to those set outin NAFTA, but these annexes explicitly state that “no pro-vision shall impose any obligation on a party regardingits immigration measures,” and they contain no marketaccess commitments. Thus, public and official attitudesin the United States with respect to labor mobility haveregressed since 2002. Until political opinion changes, itwill be close to impossible for developing countries to



in mode 4 commitments by the EU include the four thatare traditional for PTAs: traders and investors, business vis-itors, intracorporate transferees, and independent profes-sionals. The EU has negotiated relatively few PTAs withdeveloping countries that cover services. Although it hasnumerous association agreements in place with neighbor-ing Mediterranean countries (the Arab Republic of Egypt,Jordan, Morocco, the Syrian Arab Republic, Tunisia,Turkey, and others), these agreements focus on goods andhave not yet incorporated services provisions.

The EU has negotiated association agreements withMexico and Chile and has more recently finalized an eco-nomic partnership agreement (EPA) with CARIFORUM.There are no in-depth services provisions in the EUagreement with Mexico, which was concluded in March2000 when the GATS negotiations were just beginning,but the PTA with Chile is substantial. In addition to thecoverage of mode 4 in the text of the agreement, there isa specific article entitled “Movement of Natural Persons”in the EU–Chile association agreement, as well as anannex on professionals.6 In the annex, the EU specifies33 categories of professional services providers that itwill accept from Chile without numerical limit, for atime period of three months, subject to the “necessaryacademic qualification and experience.” Interestingly,Chile did not commit reciprocally to accepting any pro-fessionals from the EU.

The more recent EU–CARIFORUM economic partner-ship agreement follows a similar structure, but, in addi-tion to the usual categories of workers defined undermode 4, the EU has expanded coverage of workers to threeadditional categories important for CARIFORUM mem-bers: contractual services suppliers, independent profes-sionals, and graduate trainees. The following applies tothese three categories:7

• Contractual services suppliers. This category applies to aspecific list of activities and permits temporary entry fora cumulative period of six months. A contractual serv-ices supplier must fulfill certain requirements; the termsand conditions are set out in EU schedules for its mem-ber states.

• Independent professionals. The provisions for the CSScategory also apply to independent professionals, again,subject to EU schedules.

• Graduate trainees. Graduate trainees, a new category,are workers from CARIFORUM states who have a uni-versity degree and are temporarily transferred to theparent company or to a commercial establishment forcareer development or to obtain training in businessmethods. They may enter for a period of up to one year.

In the annex on professionals attached to the CARIFO-RUM agreement, the European Union committed to accept29 categories of professional services providers withoutnumerical limit, provided that they have a universitydegree and three years’ experience. The CARIFORUMmembers did not commit reciprocally to accept any EUprofessionals.

PTAs Negotiated by Japan

Japan has negotiated four PTAs that are of interest for thequestion of labor mobility. These are summarized inannex table 13A.5. The PTAs with Mexico and Chile arevery similar in form and content to the NAFTA-typeapproach and agreements, with a negative list of noncon-forming measures and with mode 4 treated in a chapteron temporary movement of business persons. That chap-ter defines movement for the same four categories usuallyseen in trade agreements: traders and investors, businessvisitors, intracorporate transferees, and independent pro-fessionals. Japan has set a time limit of three years forthree of these categories (all except business visitors),which is a fairly generous interpretation of length of stay.

The two more recent PTAs negotiated by Japan withcountries in Southeast Asia—those with Indonesia andthe Philippines—are notable for their innovations in cov-ering, for the first time, specific categories of nurses andhealth care workers. These PTAs have also expanded thecategories of workers in the chapter on mode 4 to include“professionals with personal contracts” (essentially, thesame as independent professionals). All these categories(except for business visitors) are allowed a stay of up tothree years. Japan has also increased the number of pro-fessional categories covered in the annex on professionalsin these two PTAs, to 14 in the case of Indonesia and10 for the Philippines. No numerical limits are placedon these professional categories, except for nurses andhealth care workers, for whom an annual quota is ineffect. For those professions, specific educational andtraining requirements are included in the agreements: ahealth degree plus two years of prior work experience andsix months of language training in Japanese. The specifi-cation of particular categories of work with annual quo-tas and training requirements is an innovative approachthat has not yet been seen in other PTAs.

PTAs Negotiated by Australia and New Zealand

Four PTAs negotiated by Australia and New Zealand arerelevant for this study (annex table 13A.6). One PTA wasnegotiated jointly by the two countries with the 10-member

Labor Mobility 283


includes the four usual categories of labor plus the addi-tional category of “installers” added by New Zealand. Thelength of stay offered by the partners to the PTA is vari-able, with Australia and New Zealand allowing the longeststay, of three or four years, respectively, for intracorporatetransferees and one year for independent professionalsand contractual services suppliers. It is notable that theASEAN members committed to much less generous dura-tions of stay for all labor categories than did their devel-oped partners. In another innovative decision, Australiaalso included “spouses” in its categories of temporarylabor permitted entry.

Comparison and Assessment of PTAs

An overall comparison of the PTAs negotiated with devel-oping countries by the United States, Canada, the EU,Japan, and Australia and New Zealand is shown in the bargraphs set out in annex figures 13A.1–13A.6. The bars indi-cate the relative magnitude of the developed economies’commitments on labor mobility. The higher the bar for aparticular agreement, the more access it provides for work-ers from its developing-country partner. The number ofcategories of workers encompassed within the chapter ontemporary entry or movement of natural persons is indi-cated, as well as the number of professionals allowed,under specified quotas or without numerical limitation,through the annex on professionals.

In general, it can be said that trade agreements con-cluded by developed countries with developing countriesfocus almost exclusively on professional services providers.Many, however, go well beyond GATS in providing accessfor a greater number of categories of professional services,through expanded numbers of covered categories orthrough the provision of unlimited access. They also oftenoffer the possibility of long-term visa renewals once profes-sionals are settled in the country. Thus, distinct progresshas been made with respect to professional services.

A few developed countries have been willing veryrecently to go beyond the expansion of access for profes-sional services providers. These include, notably, Canada,in two recent FTAs negotiated with Colombia and Peruthat extend access to the Canadian market to 50 categoriesof technicians. The innovative group also includes theEU’s EPA with CARIFORUM, which extends marketaccess to contractual services suppliers and independentprofessionals (for stays of six months) and to graduatetrainees (for stays of one year). Japan has moved to liberalizeaccess to its labor market for nurses and health care workersin its recent EPAs with Indonesia and the Philippines.Finally, both Australia and New Zealand have expanded

Association of Southeast Asian Nations (ASEAN). Threeof the four are very recent, having been signed or havingentered into force since the end of August 2008. The oldestof the four, and the one with the least ambitious provi-sions for labor mobility, is the Trans-Pacific Strategic Economic Partnership Agreement between BruneiDarussalam, Chile, New Zealand, and Singapore. Thisagreement, which entered into force in 2006, follows aNAFTA-type structure, but with lighter content. Theonly category of workers specified in the temporaryentry (or labor mobility) chapter is that of professionals,and no length of stay is specified. The annex on profes-sional services primarily sets out a best-endeavors clausefor the development of “mutually acceptable standardsand criteria for licensing and certification of professionalservice providers.” No professional categories of servicesproviders are listed, and so the annex has no marketaccess component.

In the New Zealand–China PTA that entered into forcein October 2008, the chapter on labor mobility specifiesfive categories of labor: business visitors, investors, intra-corporate transferees, contractual services supplies, and anew category of “installers.” The CSS category includesartisans with Chinese cultural expertise such as theaterartists, Mandarin language teachers, and Chinese medicalspecialists. China made no commitment with regard toprofessional service providers; New Zealand allows entryof designated professionals from China for up to threeyears. Intracorporate transferees from China are also per-mitted a three-year stay. The new category of installers isallowed a three-month stay.

The Australia–Chile PTA that entered into force inMarch 2009 follows a NAFTA-type structure, and the chap-ter on temporary entry specifies the four usual categoriesof labor. An annex on professional services does notinclude a market access component, and no numbers areattached to any category of worker. Australia allows intra-corporate transferees a stay of up to four years and con-tractual services suppliers, a stay of one year, with the pos-sibility of renewal. This recent PTA is quite original in itstreatment of spouses and accompanying family members;they are granted the right to join the worker after he or shehas been in Australia for more than one year. Dependentsand spouses of corporate executives, intracorporate trans-ferees, and contractual services suppliers from the otherparty to the agreement are allowed to enter and reside inAustralia or Chile. Moreover, the spouse is given the rightto enter, stay, and work, for a period of time equal to thatof the national.

The ASEAN–Australia–New Zealand PTA, signed inAugust 2008, contains a chapter on temporary entry, which



the categories of labor in their PTAs to include contractualservice suppliers and “installers” (for New Zealand) intheir recent agreement with ASEAN members and in theNew Zealand PTA with China. The latter also containsnovel provisions for artisans who are proficient in Chinesecultural occupations, such as theater, language, and medi-cine. Australia’s PTA with Chile covers the spouses anddependents of intracorporate transferees and contractualservices suppliers residing in the country longer than oneyear. Thus trade agreements have moved over the past twoyears beyond the purely professional categories of labor toinclude within their scope contractual services suppliers,semiprofessionals and technicians, nurses and health careworkers, and even spouses and dependents.

The trading partners that have been the most willing toopen their markets wider for foreign workers from devel-oping PTA partners have been countries that face consid-erable labor shortages. Canada has shown itself the mostgenerous in this respect, with Japan being selective andsector-specific in responding to its labor market needs.Australia has been willing to consider family dependentsas part of the labor categories defined under its mostrecent PTA. The United States and the EU have facedheavy inward migration flows, both documented andundocumented, from Latin America (in the U.S. case)and from North Africa and Eastern Europe (for the EU),and they are less willing to contractually bind greatermarket openness for foreign workers in their PTAs.8

Nonetheless, the EU did expand its coverage of labor categories in the recent EPA with CARIFORUM mem-bers. In the United States, official and public attitudeshave turned sour, and no agreements have been negoti-ated containing mode 4 coverage since 2002.

The story of labor mobility within trade agreements isstill being written; the situation continues to evolve. Cur-rently, several PTAs between developed and developingeconomies are under negotiation. The EU is negotiatingwith ASEAN, Colombia, Ecuador, India, Korea, Peru, fivecountries in Central America, and the four members of theSouthern Cone Common Market (Mercosur, MercadoComún del Sur). Canada is negotiating with the CaribbeanCommunity (CARICOM), four countries in CentralAmerica, the Dominican Republic, Jordan, Korea, Panama,and Singapore. Japan is negotiating with India and Peru,and Australia is negotiating with China, the Gulf Coopera-tion Council (GCC), Korea, and Malaysia. Only the UnitedStates is currently abstaining from further involvement inregional trade negotiations. Thus, the sample for evaluat-ing the treatment of labor mobility in PTAs will continueto expand in the coming years. Developing countries thatare able to proactively define and push their interests with

developed-country trading partners should find opportu-nities that did not exist in the past.9

PTAs Negotiated among Developing Countries

Labor mobility is not confined to PTAs involving devel-oped countries; it is also an important feature of severalSouth-South PTAs. This section presents a brief overviewof labor mobility provisions in several agreements.

The Caribbean Community (CARICOM) has had oneof the most successful experiences with liberalizing themovement of service providers at the regional level, dat-ing from the signing in 1998 of Protocol II, on “Establish-ment, Services and Capital.” The objective of the protocolis to bring the CARICOM Single Market Economy intoeffect. (For further discussion, see Stephenson 2007.)CARICOM provisions rest on two pillars: (a) facilitationof travel, common travel documents, and national treat-ment at the port of entry (Article 46 of the CARICOMSingle Market Economy treaty) and, since 2005, a com-mon passport; and (b) the free movement of skilled per-sons within the community (Articles 32, 34d, 35d, 36,36a, and 37 of the treaty). Five categories of skilled work-ers were initially identified:

• Graduates of the Universities of the West Indies,Guyana, and Suriname

• Graduates of approved universities outside the region• Media workers• Musicians• Artists• Sports persons• Workers in the tourism and entertainment industries• Any other skilled person eligible under Articles 35d and

36a of Protocol II.

Effective January 1, 2010, domestic helpers were addedto the list.10 Since 2007, discussions have been under wayon adding teachers and nurses.

A certificate of recognition must be obtained from therespective national labor ministries by those wishing tomove abroad. A six-month temporary residency permit isissued while the certificate is reviewed by the receivingcountry, after which an indefinite work and residencepermit is granted. CARICOM has recognized the impor-tance of transferability of social security benefits, butprogress on this matter has been slower than expected.

Achieving the free movement of workers is also statedas a goal of the East African Economic Community in itsprotocol on establishing a common market, expected toenter into force on July 1, 2010. Many hurdles remain,

Labor Mobility 285


of residence.14 It further provides for nondiscriminationwith respect to the right to seek and engage in employ-ment. In 2005 the conference of WAEMU/UEMOA headsof state and government approved a progressive approachtoward the implementation of freedom of movement forpersons, the right of residence, the provision of services,and the right of establishment. This suggests the adoptionof regional codes of freedoms and rights of movement, aswell as harmonization measures. The codes concern fourareas (OECD 2008):

• Right of establishment for the freedom to carry out self-employed professions

• Under equal conditions, access to higher-educationestablishments

• Establishment of a community visa for nationals ofcountries outside the WAEMU/UEMOA or ECOWASzones

• Building of control posts juxtaposed on both sides ofthe border of member countries.

In 2006 regulations were adopted on free movement andright of establishment for workers in specific professions(for example, accountants and pharmacists). Today, thecommission is working on a draft common policy in theareas of movement and stay by third-country nationals.

In Latin America, as in Southeast Asia, progress in lib-eralizing labor mobility has been slow within regionalarrangements. Mercosur members have included free-dom of movement among their integration goals. In the-ory, Mercosur nationals may currently move amongmember states, although the right to work is regulated byhost governments. Progress in liberalizing labor mobilityhas been sluggish. A Mercosur social security agreementwas signed in 1997, but many of the steps aimed at facil-itating migration within the community are taking farlonger to be implemented than planned. Much of themigration that occurs in the Mercosur region is outsideformal channels. In December 2002, Mercosur leaderssigned an Agreement on Residency for Mercosur Nation-als aimed at giving migrants “equal civil, social, cultural,and economic rights and freedoms” with the citizens ofthe Mercosur country in which they are living, “particu-larly the right to work and to carry out any legal activity.”The related Agreement on Regulating the Migration ofMercosur Citizens encouraged Mercosur governments tolegalize unauthorized nationals of Mercosur members(World Bank 2010).

In October 2003, ASEAN members raised their ambi-tions from the formation of a free trade area to the cre-ation of an ASEAN Economic Community (AEC), in the

however, on the path to full free movement. Work per-mits are not harmonized across countries; they are rela-tively restrictive and remain difficult to obtain; they aresubject to delays and administrative requirements; andrejections are numerous. Portability of social benefits isvery limited. Progress has been made on harmonizationof standards and mutual recognition for graduates, butmuch less so for technical and vocational training.

The treaty of the Economic Community of WestAfrican States (ECOWAS) requires the community toensure the removal of obstacles to the free movement ofpersons, goods, services, and capital and to guarantee theright of residence and establishment. To date, ECOWAShas signed three supplementary protocols on this subject.The first provides for the free entry of community citizensfor a period of 90 days without a visa, provided that theypossess travel documents, and also grants them the rightsof entry, residence, and establishment.11 The second pro-tocol allows community residents to reside, seek employ-ment, and engage in income-earning employment in anymember state.12 The protocol specifically refers to migrantworkers, defined as nationals of community memberstates who seek or propose to hold employment, arealready holding employment, or have in the past heldemployment in a member country. Special provisions aremade for four protocol categories: migrant workers, itin-erant workers, seasonal workers, and border workers(Mattoo and Sauvé 2010).

Despite early provisions on the free movement of persons, implementation within ECOWAS has beenslow—hampered, in particular, by the efforts of youngmember states to affirm their sovereignty. At times, slowprogress in other areas of economic integration andadverse reactions to the influx of foreign labor in periodsof recession have hindered implementation (OECD2008). In recent years, several measures have been under-taken—in particular, since 2000, in the harmonization ofpassports, as well as joint border operations by customsand migration offices.13

An interesting development took place in January 2008when ECOWAS adopted a common approach to migra-tion, clearly influenced by the European model (OECD2008). The approach consists of two parts, the first devotedto the legal framework and key principles, and the secondto a regional migration and development action plan.

For countries that are also members of the West AfricanEconomic and Monetary Union/Union Économique etMonétaire Ouest-Africaine (WAEMU/UEMOA), the treatyconfers the right of free movement of people, the right toprovide services, the right of establishment of persons car-rying out an independent or salaried activity, and the right



Declaration of the ASEAN (Bali) Concord II, which wassubsequently endorsed by summits of ASEAN leaders (Soe-sastro 2005). Nevertheless, the liberalization of labor move-ment still has a very long way to go before achievement ofthis objective. Currently, ASEAN members have madeonly very modest commitments on mode 4 in theirrespective schedules of services commitments, whichhave now undergone five rounds of negotiations underthe ASEAN Framework Agreement on Services since thatpact went into effect in January 1996. Many of the mode4 commitments go no farther than what is set out inmembers’ WTO schedules. ASEAN has, however, mademore progress with the realization of mutual recognitionagreements (MRAs) than any other regional grouping,having signed six MRAs to facilitate the movement ofprofessional services suppliers through the recognitionof their professional accreditations. These agreementscover engineering services, nursing services, architec-tural services, medical practitioners, dental practitioners,and accountancy services.

Bilateral Labor Agreements

Efforts to manage labor mobility among developingcountries at the regional level have been consequential,although progress has not matched aspirations. The mostnotable efforts have taken place in agreements aiming fora common market; these generally go beyond simplymanaging labor mobility to encompass migration dimen-sions. The examples reviewed above all concern countriesthat are regional neighbors with a significant history ofpopulation migration. It is worth highlighting someinteresting initiatives, which include coverage of recentgraduates, common migration policies, and commonpassports to facilitate border crossings.

As was noted above, the treatment of labor mobility informal PTAs has focused overwhelmingly on skilled laborcategories, with only a few recent agreements moving tocover certain types of semiskilled workers. Against thisbackground, can other vehicles be used to promote labormobility? Bilateral labor agreements (BLAs) are alterna-tives to the more legalistic and rigid PTAs and can serveboth to promote and to regulate the flow of unskilled orsemiskilled workers.

Short History

Bilateral labor agreements have provided a means foremploying seasonal and low-skilled foreign labor on atemporary basis. They allow industrial countries that needforeign labor to design labor exchange programs that steer

inward flows to specific areas of labor demand. For desti-nation countries, the primary aim is to address skill gapsin the local labor market, whether for seasonal workers(often in the agricultural sector) or low-skilled labor.Occasionally, BLAs also deal with higher-skilled workersin areas of labor shortages such as health or informationtechnology.15

Bilateral labor agreements have had an interesting his-tory. They were popular in the United States and Europein the 1960s but fell into disfavor in the 1970s and 1980s,affected by the adverse combination of inflation and highunemployment that came to be known as “stagflation.”For 22 years, beginning in 1942, the United States had abracero program to admit temporary agricultural workersfrom Mexico. Admissions under this program peaked atmore than 450,000 a year but began to shrink because ofthe enforcement of labor market regulations, combinedwith technological changes. Nonetheless, the programcontinued to admit more than 200,000 Mexican tempo-rary workers a year until it ended in 1964. In WesternEurope, temporary worker programs were peaking whenthey were ended unilaterally in 1973–74. European tem-porary worker programs differed from the Mexico–U.S.program in several important respects, including the locusof employment (nonfarm manufacturing, construction,and mining, rather than agriculture), as well as in theirpolicies toward settlement. Unlike Mexicans who filledseasonal U.S. jobs and were expected to return to Mexicoevery year, migrants in Europe filled year-round jobs andearned rights to unify their families and settle with workand residence permits.16

Several developed countries have entered into second-generation bilateral labor agreements (as of the turn ofthe millennium), although many of these were in theform of memoranda of understanding, rather than moreformal contractual arrangements. BLAs do not take anyone set form; in fact, there is such a variety of agreementsthat international organizations have developed a “Com-pendium of Good Practice Policy Elements in BilateralTemporary Labour Arrangements” as a follow-up activityto the first Global Forum on Migration and Development(GFMD), held in 2007.

BLAs have been increasing in number over recent years,but no single institution is responsible for collecting andmaintaining information on them. Neither the ILO northe IOM (International Organization for Migration) hasinformation on BLAs at the country level.17 It is thereforeextremely challenging to collect these data, and what ispresented in this section will certainly be incomplete.Although many countries have entered into bilateral laboragreements, others prefer to channel their temporary

Labor Mobility 287


agricultural workers. The approved requests are then com-municated via Canadian network contacts in Mexico to pri-vate recruitment agencies in the participating Caribbeancountries. Finding the workers to fill the required demandis then the responsibility of the countries of origin. In2000 about 7,300 Mexicans were among the 16,900 foreignfarmworkers admitted to Canada; the other workers werefrom Barbados, Colombia, Jamaica, Trinidad and Tobago,and six other Eastern Caribbean islands.19 The BLA withColombia is the result of demands by Canadian compa-nies in Alberta and Manitoba for Colombian workers inthe food-packing industry.

The BLAs concluded by Spain provide for a selectioncommittee that is made up of representatives of the par-ticipating governments and is responsible for selectingthe best-qualified workers for existing job offers and forconducting training courses that may be needed. In theseagreements for regulating labor migration flows, theSpanish authorities, through Spanish embassies in origincountries, notify authorities in the origin country of thenumber and types of workers needed. There is no setquota. Origin countries in turn notify the Spanishauthorities of the possibility of meeting this demand withtheir nationals willing to go to Spain.

Spain’s bilateral agreement with Colombia covers agri-cultural workers who are selected to work temporarily infruit harvesting in the Catalonia region. Within the frame-work of the temporary and circular labor migration modethat implements this agreement, the National TrainingInstitute in Colombia designs training programs for thelabor migrants so that, on their return to their communi-ties of origin, they can transfer the skills and know-how

labor needs through their more formal immigration chan-nels. In the United States, most temporary admission pro-grams are open to citizens of all countries. The range oftemporary visa programs includes both skilled profession-als (e.g., H-1B visas) and other kinds of temporary labor(e.g., H-2A temporary agricultural workers).

In examining the panorama of bilateral labor agree-ments that we have been able to identify for this study andthat are set out in table 13.1, it is interesting to note thatsuch agreements have now been signed by countries in allregions of the world. In the Americas, Canada has beenvery active in developing bilateral temporary worker pro-grams and has concluded agreements with Barbados,Colombia, Guatemala, Jamaica, Mexico, Trinidad andTobago, and the countries of the Eastern Caribbean. InEurope, the governments of Germany, Italy, Spain, andthe United Kingdom have actively negotiated bilaterallabor agreements with developing countries aroundthe world. In Africa, South Africa has pursued sucharrangements, mainly with neighboring countries. In Asia,China has concluded BLAs with several developed- anddeveloping-country partners.18

Terms of Coverage

The coverage of the bilateral labor agreements varies.Canada’s agreements cover exclusively the agricultural sec-tor. The Seasonal Agricultural Workers Program (SAWP)is based on bilateral memoranda of understanding and ismanaged by Human Resources and Skills DevelopmentCanada (HRSDC). Canadian employers submit requests,which have to be approved by the HRSDC, for foreign


Table 13.1. Bilateral Labor Agreements with Developing-Country Partners: Government Programs for Temporary Workers

Region and country Developing-country partners

Americas and the CaribbeanCanada Barbados, Colombia, Guatemala, Jamaica, Mexico, Organization of Eastern Caribbean States, Trinidad

and TobagoEuropeFrance MauritiusGermany Bulgaria, Croatia, Czech Republic, Poland, Romania, Slovak Republic, Slovenia, UkraineGreece Albania, BulgariaItaly Albania, Moldova, Sri Lanka, TunisiaSpain Bulgaria, Colombia, Dominican Republic, Ecuador, Mauritania, Morocco, Philippines, Romania, SenegalUnited Kingdom India, Philippines, SpainAsiaChina Australia; Japan; Jordan; Korea, Rep.; Mauritius; South Africa; Spain; United Arab EmiratesAfricaSouth Africa Botswana; Cuba; Iran, Islamic Rep.; Lesotho; Malawi; Mozambique; Swaziland; Tunisia

Source: ILO, IOM, and OSCE, “Compendium of Good Practice Policy Elements in Bilateral Temporary Labour Arrangements,” revised version, December 2, 2008. Note: ILO, International Labour Organization; IOM, International Organization for Migration; OSCE, Organization for Security and Co-operation in Europe.


acquired in Catalonia. Under this BLA, less than 10 per-cent of selected Colombian workers have failed to returnhome. In Ecuador a Migration and Control Unit was cre-ated in 2002 within the Ministry of Foreign Affairs toreceive job vacancy notices from Spanish enterprises andmatch the job offers with the most appropriate candi-dates through a large database. Spain has similar BLAprograms with Bulgaria, the Dominican Republic, Mauritania, Morocco, Romania, and Senegal. Under theSpain–Philippines BLA, nurses and other Filipino work-ers are allowed into Spain and are afforded the same pro-tections as Spanish workers.

The bilateral labor agreements signed by the UnitedKingdom with India, the Philippines, and Spain enable theUnited Kingdom to recruit registered nurses and otherhealth care professionals (physiotherapists, radiographers,occupational therapists, biomedical scientists, and otherworkers regulated by appropriate professional bodies inboth countries) for work on a temporary basis. TheU.K.–Spain agreement provides for recognition of Spanishnursing skills in the United Kingdom.

Greece has signed BLAs in the agriculture and fisheriessectors. Under the agreements with Albania and Bulgaria,Greek authorities assess the annual need for seasonalagricultural workers and grant residence and permits toworkers from these countries according to demand fromGreek employers. Under the BLA with Egypt, which cov-ers the fisheries sector, temporary labor migrants are sub-ject to specific regulations regarding the possibility ofchanging employers and the extension of their stay in thecountry, and they are eligible for the transfer of socialsecurity rights and pensions on a mutual basis.

South Africa has negotiated several bilateral agree-ments with neighboring countries in response to its grow-ing labor crisis. The Joint Initiative for Priority SkillsAcquisition (JIPSA) Act of 2004 acknowledged that par-ticular sectors require skills from outside the country.South African mining companies fought hard to keeptheir right to hire foreign contract workers, and the 2002Immigration Act was modified to accommodate this pres-sure. Bilateral agreements are focused on recruiting work-ers from Botswana, Lesotho, Malawi, Mozambique, andSwaziland to work in the mines and farms of South Africa.The share of foreigners in the mines’ workforce rose from47 percent in 1990 to 60 percent in 2000, but this share hasdeclined recently in response to efforts to hire locally.

China has negotiated several BLAs with willing partnersexperiencing labor shortages, including Australia, Japan,Jordan, Korea, Mauritius, South Africa, Spain, and theUnited Arab Emirates. These agreements are very diverse,cover a wide range of topics on labor cooperation, and list

specific numbers of recruited workers. The number of Chi-nese citizens working as temporary laborers abroad hasincreased substantially, from 63,200 in 1987 to more thanhalf a million in 2004.20

The BLAs in which China is a partner cover diverselabor sectors. Chinese labor cooperation with the UnitedArab Emirates takes place in the areas of construction,factories, medical care centers, and maritime activities.With Australia, the BLA centers on nursing and a fewother sectors, and an attempt is made to curb the exces-sive fees charged by the recruitment agencies by offeringthe alternative of government employment offices. Underthe BLA with Mauritius, Chinese workers may be recruitedonly from companies that are on an established, govern-ment-approved list. The BLA with Jordan concerns the tex-tile and construction sectors. An agreement with SouthAfrica was concluded in 2002 and was extended in 2006 tofocus on human resources development and job creationstrategies, in addition to worker recruitment. A successfulpattern for the bilateral labor agreement has been theagreement between China and Japan, under which morethan 30,000 Chinese trainees are sent to Japan every year intemporary labor (trainee) cooperation programs.21

The advantages of BLAs or temporary worker pro-grams, particularly for lower-skilled categories of work-ers, are numerous. First, and most important, is the flexi-bility they offer with respect to the management of thelabor market by the countries involved. Such agreementscan be negotiated in response to the economic cycles ofthe market.22 As is seen in the examples cited, they can betargeted to specific sectors and can even be firm based, ifnecessary. Monitoring of such agreements can be carriedout on both sides as a joint responsibility, rather thanputting the burden entirely on the destination country todetermine the legality of the worker. Guarantees can bedesigned and written into the agreements in the form ofbonds or fines for noncompliance, to encourage respectfor the provisions by private parties. Incentives canbe built in on both sides of the agreement. Workers aremore willing to respect the contract and return home if there are prospects of an opportunity (based on per-formance and need) to go back to the host country forfuture employment. Most participants in the BLAs thathave functioned to date have found that the agreementshave fulfilled the expectations of both sides.23

The disadvantage of bilateral labor agreements is that,unlike PTAs, they are single-issue instruments. This limi-tation means that developing-country partners do nothave scope within a BLA to trade their “offensive” interestsin labor mobility for the “offensive” interests of theirdeveloped-country trading partners.

Labor Mobility 289


be applied with “flexibility” when a developing country is aparty to the economic integration agreement, under theprovisions of Article V(2). In our judgment, the AppellateBody would give great weight to Article V(2) in evaluatinga BLA. In fact, we think the “flexibility” provision would bedecisive.

Third, there is the matter of negotiating history. As ourdiscussion shows, BLAs, like BITs, have been around for avery long time, predating the original GATT (signed in1947) by decades. A strong argument can be made that if theUruguay Round negotiators had meant to impose an MFNrequirement on these agreements, they would have said so invery explicit terms. After all, important economic arrange-ments would have been upset by an MFN requirement.Silence seems to indicate assent to the status quo ante.

One of the main reasons that countries enter into BLAs isthat these agreements are flexible and short term and appearto escape the long-term contractual constraints of GATS. Thelarge majority of bilateral labor agreements cover a differentcategory of worker than do the formal services agreements(PTAs or GATS); they focus on unskilled (agricultural) orlower-skilled workers, whose movements governments havenot been willing to liberalize in the context of either GATS orregional services agreements. We conclude, from this briefand sketchy review, that the Appellate Body would very likelyrespect the special status of BLAs if a claim were everbrought. As a practical matter, no WTO member has muchinterest in bringing a claim, and the possibility of litigationseems remote.

Conclusions

This chapter has examined how recent preferential tradeagreements (those concluded since 1994 between devel-oped and developing trading partners, as well as someSouth-South PTAs and labor agreements) have dealt withlabor mobility. It has shown that some of the most recentPTAs have innovated in interesting ways to promote labormobility, either by expanding the number of services sup-pliers accepted under particular categories (for example,without numerical quotas) or by creating new space forspecifically defined categories of labor, such as techni-cians, nurses and health care workers, and sporting andcultural occupations. To date, however, all but a very fewPTAs that cover services focus on professional servicessuppliers. A new generation of less formal temporaryworker programs is paying more attention to the needs oflower-skilled and semiskilled temporary workers. Someof the regional integration groupings among developingcountries in Africa, Asia, and Latin America and theCaribbean are making progress toward the opening of

Multilateral and MFN Considerations

It must be recognized that bilateral labor agreements repre-sent an important derogation from the most favorednation (MFN) principle that is the core of the world trad-ing system. The same is true of bilateral investment treaties(BITs) and double-taxation treaties (DTTs), of which thereare many thousands in the world today. None of these pre-tend to treat all countries on equal terms; partners arefavored over nonpartners. Of course PTAs also discrimi-nate between partners and outsiders, and those inside thePTA receive more favorable treatment than those outside.From the perspective of true multilateralism (“no discrim-ination is the first-best policy”), these various forms ofbilateral and regional agreements clearly occupy a second-best world. The tension that characterizes negotiations overtrade liberalization, as well as over investment and labormobility, is between a first-best multilateral approach,which may be stalled because of lack of agreement amongcountries worldwide, and a second-best regional or bilateralapproach that achieves liberalization between the partnersbut creates discrimination against the rest of the world. Alarge and robust literature has developed around the debateas to which approach will engender the most liberalizationand the greatest gains over the horizon of a decade orlonger. We will not rehearse the arguments here, since theyare familiar to most readers; we simply observe that theissue is certainly not settled. For the past decade, however,most governments have been “voting with their feet” byplacing more emphasis on bilateral and regional agree-ments. This trend seems very likely to continue.

What about the consistency of BLAs with GATS? Thisquestion has not been litigated in the WTO and is notlikely to be litigated any time soon, so a definitive answercannot be given. Three considerations, however, wouldprobably have weight in the WTO Appellate Body’s rea-soning, if a nonparty to a BLA did claim that its GATSrights to labor mobility (mode 4) were violated by thebilateral agreement.

First, GATS Article II(1) establishes the MFN principlefor services, including mode 4. Under Article II(2), MFNcan be waived for a BLA, as for any other GATT or GATSobligation, but this requires a favorable vote by three-fourths of WTO members. In any event, there are no extantwaivers for BLAs or BITs.24

Second, under GATS Article V(1), two or more coun-tries can enter into an economic integration agreement toliberalize trade in services and thereby avoid the MFNrequirement. The agreement should have “substantial sec-toral coverage” and should eliminate “substantially all dis-crimination” between the parties. These conditions are to



labor markets at the regional level to all categories ofworkers, both for temporary movement and for perma-nent settlement. Members of these groupings appear will-ing to go farther in their ultimate objectives than is thecase under the North-South PTAs, where the norm is tocover prescribed and limited, although often expanded,categories of workers.

Thus, while the latest steps are positive and encouragingfor developing countries, they leave much work for futurenegotiators. In our view, patience should be the watchwordof negotiators based in developing countries. They shouldtake heart and guidance from the long experience of devel-oped and developing countries in crafting the liberaliza-tion of trade in textiles and clothing. This was a supersensi-tive industry as early as the late 1950s, when theEisenhower administration in the United States negotiatedthe first restraint agreement with Japan, and it remainedsensitive for the next 50 years. Eisenhower’s accord withJapan was followed by the Short-Term Cotton Agreementand the Long-Term Cotton Agreement in the Kennedy andJohnson administrations, and then three generations of theMultifibre Arrangement (MFA) under GATT auspices.

The complexity of bilateral textiles and clothing quotasunder these agreements was truly bewildering and, from aneconomist’s viewpoint, highly distortive. But within thiscomplex framework, over the span of five decades, trade inthe sector was greatly liberalized and grew enormously.The secret, if there was a secret, was that negotiators ofgood will, representing both developed and developingcountries, discovered niches of textiles and clothingtrade where the political costs of further liberalization,combined with suitable safeguard mechanisms, were notinsurmountable. At every stage of this long process, theeconomic gains from liberalization were enormous; the“magic,” if there was any magic, was to focus attention onproducts and mechanisms that did not encounter over-whelming resistance in the developed countries. We thinkthe same approach commends itself to labor mobility nego-tiations—a long, persistent, and patient search for niches inthe labor markets of developed countries where greaterentry of migrants is not only tolerated but welcomed.

On the basis of this overview, we offer four sets of rec-ommendations.

1. Concerning professional workersWhen developing countries are able to define their inter-ests well and are willing and able to pursue bilateral tradeagreements with the major developed trading partnersreviewed in this study (other than the United States, atpresent), they should be able to obtain expanded marketaccess. Labor markets worth exploring are opportunities

for firms and individuals that offer unique cultural talentsor specialized skills, as well as for some independent pro-fessionals, and geographic or occupational niches of theindustrial developed economies that suffer from laborshortages. If developing countries wish to promote exportsof services providers in the health services, this is certainlyan area that offers a large potential for expansion. For thismarket, it might be advisable to develop local training pro-grams for the specific skills required in the target market,in the way that the Philippines has done and Indonesia iscurrently doing.

2. Concerning semiskilled and lower-skilled workersIn the case of workers with lower skill levels and less formaleducational training, the best vehicle for promoting greaterlabor mobility is not formal PTAs but the more flexibleinstrument of temporary worker programs (TWPs). Theseprograms can be designed to promote circular migrationin a way that benefits the labor-sending and labor-receiv-ing countries, as well as the workers themselves. TWPs areextremely flexible in both design and execution and allowthe parties involved to design the clauses covering length ofstay, nature and place of employment, and appropriateguarantees. They also offer governments the possibility ofadjusting in a responsive manner to the cycles of theirdomestic labor markets. Such agreements must elicit thepositive involvement of parties on both sides, making this aframework with buy-in, where all parties to the agreementhave an interest in seeing it succeed. Although these agree-ments have been successfully promoted so far by only ahandful of countries, primarily China and the Philippines,there is tremendous scope for their further application inthe world economy.

3. For developing-country governments and negotiatorsDeveloping country governments and negotiators shouldbear six precepts in mind:

• Developing-country negotiators should approach thediscussions of labor mobility with a positive attitudeand should emphasize the gains to the destinationcountry. The economic gains are invariably large, andthe political costs are often exaggerated, so it is usefulfor negotiators from developing countries to researchparticular labor markets and lay the facts on the table.

• To better serve their negotiators, developing-countrygovernments should conduct in-depth research on thelabor markets of potential destination countries withthe aim of discovering promising niches. This willrequire the services of specialized officers or contractorsworking in the destination countries.

Labor Mobility 291


respectful people. When adverse incidents happen, asthey will, the government of the developing countryshould cooperate as appropriate, through revocation ofvisas and other measures.

4. For developed-country governments and negotiatorsLike developing countries, developed countries shouldproactively search for labor market niches where addi-tional temporary workers will become valued members ofthe workforce and the community. Developed-countryofficials must not surrender to arguments that the labormarket is an undifferentiated mass, or succumb to the anti-immigrant voices of a vocal minority. Theyshould hammer home the distinction between permanentimmigration, which remains under sovereign control,and temporary workers who are subject to negotiatedagreements. They should seek to build flexible responsesnot only into TWPs but also into the quota and timeclauses of PTAs. Most important, they should put someeffort into seeking out and conveying positive messagesabout the contributions and accomplishments of tempo-rary workers.

• Developing-country specialists should work with edu-cational and credentialing authorities in the developedcountries to lay the groundwork for mutual recognitionagreements for the benefit of their independent profes-sionals and other highly skilled workers.

• When multinational corporations seek to expand theiroperations abroad, whether in a developed or a develop-ing country, government negotiators should team up withthe corporations to ensure agreement on the requisitenumber of visas for intracorporate transferees and con-tractual services suppliers to support the new operation.This needs to be done whether or not a PTA is in place.

• Developing-country negotiators should seek agreementon the status of mode 4 workers, meaning their rights asto visas, working conditions, social security contribu-tions, unemployment compensation, and ability to remitfunds. To some extent, these matters are covered inTWPs, but important elements are often not addressed.

• Above all, senior officials in the developing countrymust attend to the “image” of their migrants abroad—doing whatever is possible to ensure that their migrantsconvey an impression of hard-working, law-abiding,


Annex Table 13A.1. Quantitative Estimates of Gains from Increased Labor Mobility

AuthorsRegioncovered Assumption or situation

Borjas (1999) UnitedStates

U.S. 1980–2000 immigrationwave, with immigrantsrepresenting roughly 10 percent of the U.S. workforce

Large redistributive effect: returnto capital, +2 percent of GDP;labor wages, –1.9 percent of GDPa

Small net gains for natives:US$10 billion a year (0.1percent of U.S. GDP), or roughly5 percent of average economicgrowth over past 20 years

Hamilton and Whalley (1984)

World Elimination of all restrictions on labor mobility (1977 data)

60.1–204.6 percent of 1977 world GDP in 1977b

Iregui (1999) World Elimination of all restrictionson labor mobility (between 37 and 53 percent of the laborendowment of developing regions migrates)

Nonsegmented labor market:15–67 percent of world GDP

Segmented labor market (skilledversus unskilled): 13–59 percent of world GDP

If only skilled labor migrates:3–11 percent of world GDP

Moses and Letnes (2004)

World Elimination of all restrictions on labor mobility (1977 and 1998 data)

For 1977, with 100 percentelimination of wage differential:US$0.34 trillion–US 11.27 trillion(1977 dollars) (more probably,US$0.58 trillion); 4.3–111.6percent of 1977 world GDP (more probably, 7.5 percent of1977 world GDP)

For 1977, with 10 percentelimination of wage differential:22 percent of total potentialgain; wages, +4.1 percent inpoorest countries; +3.3 percentin middle-income countries; –2.5 percent in richest countries;return to capital, –8.3 percent inpoorest countries; –6.9 percentin middle-income countries;+5.7 percent in richest countries

For 1998, with 100 percentelimination of wage differential:US$1.97 trillion–US$55.04trillion (1998 dollars) (moreprobably, US$3.4 trillion);5.6–155 percent of 1998 worldGDP (more probably, 9.6percent of 1998 world GDP)

For 1998, with 10 percentelimination of wage differential:23 percent of total potential gain;wages, +11.4 percent in poorestcountries; +2.1 percent inmiddle-income countries; –3.1 percent in richest countries;return to capital, –21.0 percent in poorest countries; –4.4 percentin middle-income countries; +7.2 percent in richest countries

(continued next page)

Annex

Gains


Labor Mobility 293

Walmsley and Winters (2002)

World Increase in migration fromdeveloping countries to high-income countries sufficient to increase labor force in the host countries by 3 percent in 2002

Total: +0.6 percent of world GDP(US$156 billion in 2002, or1.5 times the expected gains from liberalization of all remaining goods)

Movement of unskilled workers(accounting for most of the gains):+US$110 billion versus +US$46billion for the movement of skilledworkers

Migrants’ welfare: +US$171 billion(+US$73 billion in the United States, +US$25 billion in Japan,+US$68 billion in the EU)

Resident welfare: net, –US$15 billion; developing countries, insome cases, gain if remittances are high (+US$16 billion in India),but most lose (–US$7 billion inBrazil); developed countries, smallgains (+US$3.9 billion in EU)

Change in real wages of unskilledworkers: increase in developingcountries (+0.7 percent inIndia); decrease in developedcountries (–0.6 percent in theUnited States)

Change in real wages of skilledworkers: dramatic increase indeveloping countries (+4.5 percent in Mexico); decrease in developed countries (–0.8 percent in the United States)

Change in rental price of capital:decrease in developingcountries (–0.4 percent inMexico); increase in developedcountries (+0.8 percent in theUnited States)

World Bank (2006), 31

World Increase in migration fromdeveloping countries to high-income countries sufficient to increase the labor force in the host countries by 3 percent by 2025 (revision of Walmsley and Winters 2002)

+0.6 percent of world GDP, (US$356 billion in 2025);+0.4 percent of developed-country GDP; +1.8 percent ofdeveloping-country GDP (including migrants’ income)

Source: Studies listed under “Authors”; see the bibliography for details. Note: EU, European Union; GDP, gross domestic product. a. Hatton and Williamson (1998) find similar results on wages when studying the 1870–1910 migration wage in the United States; they estimate that U.S.wages in 1910 would have been 11 to 14 percent higher in the absence of immigration after 1870. b. The large differences in estimates, both within and between studies, can be explained by differences in modeling frameworks (partial versus generalequilibrium), production elasticities, productivity, cost of movement, or workforce size.

Annex Table 13A.1. (continued)

AuthorsRegioncovered Assumption or situation


Annex Table 13A.2. Agreements between the United States and Developing Countries

Provision U.S.–Singapore U.S.–Chile U.S.–Morocco CAFTA–DR U.S.–Peru

Entry into force January 1, 2004 January 1, 2004 January 1, 2006 March 1, 2006 February 1, 2009Chapter on trade in

servicesCh. VIII Ch. 11 Ch. 11 Ch. 11 Ch. 11

Treatment of foreign servicesNational treatment Yes (Art. 8.3) Yes (Art. 11.2) Yes (Art. 11.2) Yes (Art. 11.2) Yes (Art. 11.2)Most favored nation Yes (Art. 8.4) Yes (Art. 11.3) Yes (Art. 11.3) Yes (Art. 11.3) Yes (Art. 11.3)Local presence

requiredNo (Art. 8.6) No (Art. 11.4) No (Art. 11.5) No (Art. 11.4) No (Art. 11.4)

Provisions on mode 4 Chapter Ch. 11 Ch. 14 None None None Committee Yes (Art. 11.7) Yes (Art. 14.5) Joint committee to

review theimplementation ofthe annex onprofessionals

Commission to review theimplementation of the annex onprofessionals

—

Dispute settlement Yes (Art. 11.8) Yes (Art. 14.6) — — —Transparency of

regulation Yes (Art. 11.5) Yes (Art. 14.4) — — —

Gains



Side letters Yes (professionalsmust comply withcertain labor andimmigration lawsand have anemployer in theUnited States)

Yes (professionals will obtain visathrough the U.S. H-1B program)

None Yes (“No provisionshall impose anyobligation on aparty regarding itsimmigrationmeasures”)

—

Worker categoriescovered

Investors, traders,intracorporatetransferees,professionals

Investors, traders,intracorporatetransferees,professionals

— — —

Specification of lengthof stay

None None — — —

Provisions on professionalsAnnex on

professionalsApp. 11.A.2 Annex 14.3.D Annex 11.B Annex 11.9 Annex 11.B

Number ofprofessionalcategoriescovered

2 (disaster relief claims adjuster,managementconsultant)

4 (disaster relief claims adjuster,managementconsultant,agriculturalmanager, physicaltherapist)

0 (pledge to work on)



Specified quotas Singapore: nonumerical limit

United States: 5,400

Chile: no numericallimit

United States: 1,400

— — —

Postsecondary degree required

Yes: 4 years or more Yes: 4 years or more — — —

Specification of length of stay

None None — — —

Source: Authors’ compilation.Note: — = no provisions; CAFTA–DR, Dominican Republic–Central America Free Trade Agreement.


Provision U.S.–Singapore U.S.–Chile U.S.–Morocco CAFTA–DR U.S.–Peru

Annex Table 13A.3. Agreements between Canada and Developing Countries

Provision NAFTA Canada–Chile Canada–Colombia Canada–Peru

Entry into force January 1, 1994 July 5, 1997 Signed November 21, 2008

January 1, 2009

Chapter on trade inservices

Ch. 12 Ch. H Ch. 9 Ch. 9

Treatment of foreign servicesNational treatment Yes (Art. 1202) Yes (Art. H-02) Yes (Art. 902) Yes (Art. 903)Most favored nation Yes (Art. 1203) Yes (Art. H-03) Yes (Art. 903) Yes (Art. 904)Local presence required No (Art. 1205) No (Art. H-04) No (Art. 905) No (Art. 907)Provisions on mode 4Chapter Art. 16 Ch. K Ch. 12 Ch. 12Side letters None None None NoneCommittee Yes (Art. 1605) Yes (Annex K-05)Dispute settlement Yes (Art. 1606) Yes (Art. K-06) Yes (Art. 1206) Yes (Art. 1206)Transparency of

regulationYes (Art. 1604) Yes (Art. K-04) Yes (Art. 1204) Yes (Art. 1204)

Worker categories covered

Investors, traders,intracorporatetransferees, professionals

Investors, traders,intracorporatetransferees, professionals

Investors, traders,intracorporate transferees, professionals, technicians, spouses

Investors, traders,intracorporatetransferees,professionals, technicians



Labor Mobility 295


None None None Peru: investors, 1 year;traders, 90 days;intracorporatetransferees, 1 year;professionals, 1 year;technicians, 1 year

Canada: investors, 1 year;traders, 1 year;intracorporatetransferees, 3 years;professionals, 1 year;technicians, 1 year

Provisions on professionalsAnnex on professionals App. 1603.D.1 App. K-03.IV.1 App. 1203.D App. 1203.DNumber of professional

categories covered63 (accountant,

architect, medicalprofessional, scientist,teacher, others)

72 (accountant, architect, medicalprofessional, scientist, teacher,others)

All categories of professionals except health, sports, art,education, legal, andmanagement services; 50 categories of technicians (mechanical and avionics technician,construction inspector, food and beveragesupervisor, textiles specialist, electrician,plumber, oil and gas well driller, chef, others)

All categories excepthealth, sports, art,education, legal, andmanagement services;50 technicians(mechanical and avionicstechnician, constructioninspector, food andbeverage supervisor,textiles processing,electrician, plumber, oiland gas well driller, chef,others)

Specified quotas No numerical limit except for the United States: 5,500

No numerical limit No numerical limit No numerical limit

Postsecondary degreerequired

Yes: 4 years or more Yes: 4 years or more Yes: professionals, 4 years;technicians, 2 years

Yes: professionals, 4 years;technicians, 1 year


None None None 1 year

Source: Authors’ compilation.Note: NAFTA, North American Free Trade Agreement.


Provision NAFTA Canada–Chile Canada–Colombia Canada–Peru

Annex Table 13A.4. Agreements between the European Union and Developing Countries

Provision EU–Chile EU–CARIFORUM EU–Turkey EU–Morocco

Entry into force February 1, 2003 December 29, 2008 December 31, 1995

March 18, 2000

Chapter on trade inservices

Title III, Ch. I Pt. II, Title II, Ch. 3 None Title III (pledge to workon)

Treatment of foreign servicesNational treatment Yes (Art. 98) Yes (Art. 77) — —Most favored nation No Yes (Art. 79) — —Local presence required No (Art. 97) — — —Provisions on mode 4Chapter Art. 101 Pt. II, Title II, Ch. 4 None NoneSide letters None None — —Committee Yes (Art. 100) Yes (Art. 85) — —




Dispute settlement None Pledge to create one (Art. 87)

— —

Transparency of regulation

Yes (Art. 105) Yes (Art. 86) — —


Investors, intracorporatetransferees, businesssellers, professionals

Investors, intracorporatetransferees, businesssellers, professionals,graduate trainees

— —


EU: professionals, 3 months

Investors, 90 days;intracorporate transferees, 3 years;business sellers, 90 days; independentprofessionals,contractual servicessuppliers, 6 months;graduate trainees, 1 year

— —

Provisions on professionalsAnnex on professionals Annex VII Annex IV None NoneNumber of professional

categories coveredEU: 33 (engineer,

accounting,construction, mining, computer, legal services, others)

Chile: 0

EU: 29 (architectural, legal,accounting, engineering,computer, managementservices)

CARIFORUM: 0

— —

Specified quotas No numerical limit No numerical limit — —Postsecondary degree

required“Necessary academic

qualification andexperience”

University degree + 3 yearsexperience

— —


EU: 3 months EU: 6 months — —

Source: Authors’ compilation.Note: — = no provisions; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EU, European Union.

Annex Table 13A.5. Agreements between Japan and Developing Countries

Provision Japan–Mexico Japan–Chile Japan–Indonesia Japan–Philippines

Entry into force April 1, 2005 Signed March 27, 2007 July 7, 2008 December 11, 2008Chapter on trade in

servicesCh. 8 Ch. 9 Ch. 6 Ch. 7

Treatment of foreign servicesNational treatment Yes (Art. 98) Yes (Art. 107) Yes (Art. 79) Yes (Art. 73)Most favored nation Yes (Art. 99) Yes (Art. 108) Yes (Art. 82) Yes (Art. 76)Local presence required No (Art. 100) No (Art. 109) No (Art. 78) No (Art. 72)Provisions on mode 4Chapter Ch. 10 Ch. 11 Ch. 7 Ch. 9Side letters None None None None Committee Yes (Art. 117) None Yes (Art. 96) Yes (Art. 113)Dispute settlement Yes (Art. 118) Yes (Art. 133) Yes (Ch. 14) Yes (Ch. 15)Transparency of

regulation 4Yes (Art. 116) Yes (Art. 132) Yes (Art. 95) Yes (Art. 111)


Investors, business visitors, intracorporatetransferees, professionals

Investors, business visitors, intracorporatetransferees, professionals

Investors, business visitors, Intracorporatetransferees,professionals,professionals with“personal contracts,”nurses and care workers

Investors, business visitors,intracorporatetransferees,professionals,professionals with“personal contracts,”nurses and care workers



Provision EU–Chile EU–CARIFORUM EU–Turkey EU–Morocco


Labor Mobility 297


Japan: business visitors, 90 days; othercategories, 3 years

Mexico: business visitors, 30 days; othercategories, 1 year


Chile: business visitors, 30 days; othercategories, 1 year


Indonesia: business visitors, 60 days; othercategories, 1 year


Philippines: businessvisitors, 59 days; othercategories, 1 year

Provisions on professionalsAnnex on professionals Annex 10 Annex 13 Annex 10 Annex 8Number of professional

categories coveredJapan: 2 (engineer,

specialist in humanities orinternational services)

Mexico: 42 (accountant,engineer, lawyer,scientist, nurse, others)

Japan: 2 (engineer,specialist in humanities orinternational services)

Chile: 41 (accountant,engineer, lawyer,scientist, nurse, others)

Japan: 14 (legal andaccounting services,engineer, specialist inhumanities orinternational services,nurse, health careworker)

Indonesia: 4 (mechanicaland electrical engineer,nurse, health careworker)

Japan: 10 (legal andaccounting services,engineer, specialist inhumanities orinternational services’Japanese universitygraduate nurse, healthcare worker)

Philippines: 4 (mechanicaland electrical engineer,nurse, health careworker)

Specified quotas No numerical limit No numerical limit No numerical limit exceptfor nurses and healthcare workers

—


4 years or more 4 years or more Professionals, 4 years;nurses and health careworkers, public healthdegree + 2 years workexperience + 6 monthslanguage training

Professionals, 4 years;nurses and health careworkers, national healthdegree + 3 years workexperience + 6 monthsof training in the hostcountry to pass the host-country certification exam


Japan: 3 yearsMexico: 1 year

Japan: 3 yearsChile: 1 year

Japan: 3 yearsIndonesia: 1 year

Japan: 3 yearsPhilippines: 1 year

Source: Authors’ compilation.Note: — = no provisions.


Provision Japan–Mexico Japan–Chile Japan–Indonesia Japan–Philippines

Annex Table 13A.6. Agreements between Australia and New Zealand and Developing Countries

Provision

Trans-Pacific SEP(Brunei Darussalam,Chile, New Zealand,

Singapore)ASEAN–Australia–New Zealand New Zealand–China Australia–Chile

Entry into force May 28, 2006 Signed August 28, 2008 October 1, 2008 March 6, 2009Chapter on trade in

servicesCh. 12 Ch. 8 Ch. 9 Ch. 9

Treatment of workersNational treatment Yes (Art. 12-4) Yes (Ch. 8, Art. 5) Yes (Art. 106) Yes (Art. 9-3)Most favored nation Yes (Art. 12-3) Yes (Ch. 8, Art. 7) Yes (Art. 107) Yes (Art. 9-4)Local presence required No (Art. 12-7) No (Ch. 8, Art. 4) No (Art. 108) No (Art. 9-5)Provisions on mode 4Chapter Art. 12-11 Ch. 9 Ch. 10 Ch. 13Side letters None None None NoneCommittee None None Yes (Art. 133) Yes (Art. 13-6)Dispute settlement None Yes (Ch. 9, Art. 9) Yes (Art. 134) Yes (Art. 13-7)Transparency of

regulation None Yes (Ch. 9, Art. 8) Yes (Art. 131) Yes (Art. 13-5)




Worker categoriescovered

Professionals Business visitors, investors,intracorporate transferees,contractual servicessuppliers, installers (to install purchasedmachinery—New Zealandonly), spouses

Business visitors, investors,intracorporate transferees,contractual servicessuppliers, installers

Business visitors,investors,intracorporatetransferees,contractual servicessuppliers, relatives


No Indonesia: business visitors, 60 days; investors, 60 days; others, 2 years

Australia: intracorporatetransferees, 4 years;investors, 2 years; businessvisitors, 6 months;professionals, 12 months

New Zealand: business visitors, 3 months;investors, 3 months;intracorporate transferees, 3 years; installers, 3 months; professionals, 12 months

Philippines: business visitors,59 days; others, 1 year

Vietnam: intracorporatetransferees, 3 years; others, 90 days

China: business visitors, 6 months; investors, 6 months; intracorporatetransferees, 3 years;professionals, —; installers,3 months

New Zealand: business visitors, 3 months;investors, 3 months;intracorporate transferees, 3 years; professionals, 3 years; installers, 3 months

Australia: business visitor, 1 year;investors, 90 days;intracorporatetransferees, 4 years;professionals, 1 year

Chile: —

Provisions on professionalsAnnex on professionals Art. 12-11 Annex 4 Annexes 10 and 11 Annex 13-ANumber of professional

categories coveredPledge to “work on”

6 categories (engineer, architect,geologist,geophysicist, planner, accountant)

Australia: 0New Zealand: 33

(engineering, legal,taxation, veterinary,computer, translationservices)

Indonesia: 13 (legal, tourism,restaurant, engineering,computer, R&D,maintenance services)

Philippines: all persons “whooccupy a technical,advisory, or supervisoryposition”

Vietnam: 2 (computer andengineering services)

Singapore: 0

China: 5 (education, medical, translation, hotel, computer)

New Zealand: 6 (traditionalChinese medicine, Chinesechef, Mandarin teachingaide, martial arts coach,tour guide, skilled worker“in category identified asbeing in shortage”)

“Subject to nationalcriteria”

Specified quotas — Entry subject to national rules

China: no numerical limitNew Zealand: traditional

Chinese medicine, 200; Chinese chefs, 200; Mandarin teachingaides, 150; martial artscoaches, 150; tour guides, 100; skilled workers “in category inshortage,” 1,000

—


Provision





Labor Mobility 299


New Zealand: 3 years or more + 6 years experience

Indonesia: “high qualification”

Philippines: “knowledge at an advanced level”

Vietnam: “university degree” + 5 yearsexperience

China: “appropriate education level” + 2 years experience

New Zealand:“appropriate educationlevel” + experience

—


— Vietnam: 90 daysNew Zealand: 1 yearPhilippines: 1 yearIndonesia: 2 years

China: —New Zealand: 3 years

Chile: —Australia: 1 year

Source: Authors’ compilation.Note: — = no provisions; ASEAN, Association of Southeast Asian Nations; R&D, research and development; SEP, Strategic Economic Partnership.


Provision





quotas number of technician categories covered

number of professional categories covered mode 4 categories covered

annex on mode 4 chapter on services

0

20

40

60

80

100 10

5.42 4

1.4

63

20 20 20

5 5 55 5 5 5 5 5

120

NAFTA (1

994)

U.S.–S

ingap

ore

(200

4)

U.S.–C

hile

(200

4)

U.S.–M

oroc

co

(200

6)

CAFTA–D

R

(200

6)

U.S.–P

eru

(200

6)

Annex Figure 13A.1. Provisions on Mode 4 in PTAs between the United States and Developing Countries

Source: Authors.Note: CAFTA–DR, Dominican Republic–Central America Free Trade Agreement; NAFTA, North American Free Trade Agreement; PTA, preferential tradeagreement. Mode 4 refers to the movement of natural persons to supply services. The height of the bars indicates the degree of access that PTA provides forworkers from developing-country partners. Values are assigned to each component of access as follows (not all components may be applicable to aparticular agreement): Chapter on services in the PTA? If yes, 5 points.Annex on mode 4 service supply? If yes, 5 points.Mode 4 categories covered. Number of categories (shown on the bars) is multiplied by 5 to yield total points.Number of professional categories covered. Shown on the bars.Number of technician categories covered. Shown on the bars.Quotas. If uncapped, 10 points. Otherwise, shown as the total number of workers allowed under the quota, in thousands.


Labor Mobility 301

0

50

100

150

200

1010

6372

20 20

55

55

10

30

94(all but 6)

50

55

10

50

94(all but 6)

25

55

NAFTA (1

994)

Canad

a–Chil

e

(199

7)

Canad

a–Colo

mbia

(200

8)

Canad

a–Pe

ru

(200

8)




Annex Figure 13A.2. Provisions on Mode 4 in PTAs between Canada and Developing Countries

Source: Authors.Note: NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement. For the method of deriving the values for the bars, see the note toannex figure 13A.1.



0

10

2025

5555 5

20

pledge

uncapped uncapped

30

40

50

60

70

80

10

EU–Morocco (2000) EU–Chile (2003) EU–CARIFORUM (2008)

10

2933




Annex Figure 13A.3. Provisions on Mode 4 in PTAs between the European Union (EU) and Developing Countries

Source: Authors.Note: CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; PTA, preferential trade agreement. For the method of deriving thevalues for the bars, see the note to annex figure 13A.1.

0

20

40

60

80

100

Japan–Mexico(2005)

Japan–Chile(2007)

Japan–Indonesia(2008)

Japan–Philippines(2008)

10

44

20

55

55

55

55

2030 30

4318

14

10

1010




Annex Figure 13A.4. Provisions on Mode 4 in PTAs between Japan and Developing Countries

Source: Authors.Note: PTA, preferential trade agreement. For the method of deriving the values for the bars, see the note to annex figure 13A.1.


Labor Mobility 303

0

20

40

60

80

100

pledge

nationalrules

nationalrules

China:uncapped

New Zealand:1.8

Tran

s-Pac

ific

SEP (

2006

)

ASEAN–A

ustra

lia–

New Z

ealan

d (2

008)

New Z

ealan

d–

China (

2008

)

Austra

lia–

Chile (

2009

)

6

5

3025 25

55

55

55

55

58

11

10




Annex Figure 13A.5. Provisions on Mode 4 in PTAs between Australia and New Zealand and Developing Countries

Source: Authors.Note: ASEAN, Association of Southeast Asian Nations; PTA, preferential trade agreement; SEP, Strategic Economic Partnership. For the method of deriving thevalues for the bars, see the note to annex figure 13A.1.


the round resumed in 2007. A “signaling conference” was held at therequest of interested ministers in July 2008, but it did not elicit muchenthusiasm.

3. For an earlier discussion examining treatment of mode 4 in PTAs,see Nielson (2003).

4. The only exception is the FTA with Australia. No market accessprovisions for labor mobility were included in the text itself, but a side let-ter was added after the conclusion of the negotiations that allowed anannual quota of 10,500 Australian professionals to enter the U.S. market.This was done in 2002, and it proved to be the final straw for members ofthe U.S. Congress.

5. Sáez (2009) explains that within the EU, issues concerning trade inservices do not fall exclusively under the competence of the communitybecause they go beyond Articles 113 and 238 of the treaty that accordssupranational treaty-making powers to the community on behalf of allthe member states. Thus, implementation of the services provisions andobligations of a trade agreement must be approved by each EU memberstate in accordance with domestic laws.

6. Article 101, on movement of natural persons, contains only a review requirement: “Two years after the entry into force of this

Notes

This chapter is a modified version of a chapter in Cattaneo et al. (2010).Thibaud Delourme, a student at the Maxwell School of Syracuse Univer-sity, assisted with the research.

1. Carzaniga (2008, 478) points out that foreigners working for ahost-country company on a contractual basis as independent servicessuppliers (ISS) are covered by GATS (and, in general, by trade agreementsthat include mode 4), whereas they would not be covered if they wereemployees of the company. What distinguishes their situation is the typeof payment received: the foreign employee receives a domestic currencywage from the company in the host country, whereas the ISS is paid a feeand the contractual services supplier is paid foreign wages.

2. By December 2005, the total number of services offers had reached69, involving 93 WTO members. Of the 69, 31 were revised offers. Therehas been very little change since; the number of initial offers has increasedonly to 71. Thus, one-third of WTO members have not put any initialoffer for services forward since the beginning of the negotiations. TheDoha Round was suspended in July 2006, without any revised offers hav-ing been submitted. Services negotiations were not actively taken up when


0

20

40

60

80

100

120

140

160

180

200

Canada

NAFTA

Chile

Colombia

Peru

United States

Singap

oreChile

Moro

cco

CAFTA–DR

Peru

Japan

Mex

icoChile

Indones

ia

Philip

pines

Australia and New Zealand

TPSE

P

ASEAN

China (NZ)

Chile (A

us)

European Union

Moro

cco

Chile

CARIFORU

M




Annex Figure 13A.6. Provisions on Mode 4 in PTAs between Developed and Developing Countries

Source: Authors.Note: ASEAN, Association of Southeast Asian Nations; CAFTA–DR, –Dominican Republic–Central America Free Trade Agreement; CARIFORUM, CaribbeanForum of African, Caribbean, and Pacific (ACP) States; NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement; TPSEP, Trans-PacificStrategic Economic Partnership.


Agreement, the Parties shall review the rules and conditions applicable tomovement of natural persons (mode 4) with a view to achieving furtherliberalisation.”

7. It should be mentioned that several EU member states haveattached economic needs tests (ENTs) to their commitments on mode4 entry. Actual access provided, even under the expanded commitments,will depend on how these tests are interpreted and applied in practice.No definitions were supplied with the ENT entries, and some are appliedquite restrictively.

8. Indeed, a very recent PTA between the members of the EuropeanFree Trade Association (EFTA)—Iceland, Liechtenstein, Norway, andSwitzerland—and Colombia, signed on November 25, 2008, does noteven include an annex on professional service suppliers, and the body ofthe agreement contains no mention of the movement of natural persons,other than their definition as mode 4. EFTA members are not facing laborshortages and, in the current hostile economic climate, did not feel anypressure to include liberalization of labor mobility in their agreementwith Colombia. See http://www.sice.oas.org.

9. This approach is suggested by a paper by Henry Gao, “Report onChina’s Export Interests in Services in Australia,” 2008, which was providedto the authors. The strategy calls for collecting information on the labormarket and services export structure of the potential or current negotiatingpartner, carrying out surveys to identify those categories of labor with thegreatest potential for expansion following the removal of trade-restrictivebarriers, and building on achievements of previous PTAs in the area.

10. Two members, Antigua and Barbuda, and Belize, were allowed afive-year grace period to study the impact of free mobility for domestichelpers before adding them to their list.

11. Protocol A/P.1/5/79, relating to free movement of persons, resi-dence, and establishment, in application of Article 27 of the treaty estab-lishing ECOWAS. The successive texts complementing the free movementregime are Supplementary Protocol A/SP.1/7/85, on the code of conductfor the implementation of Protocol A/P.1/5/79; Supplementary ProtocolA/SP.1/7/86, on the second phase (right of residence) of the aforemen-tioned protocol; Supplementary Protocol A/SP.1/6/89, amending andcomplementing the provisions of Article 7 of the aforementioned proto-col; and Supplementary Protocol A/SP.2/5/90, on the implementation ofthe third phase (right of establishment) of the aforementioned protocol.

12. Supplementary Protocol A/SP.1/7/86. 13. Eight countries use the regional passport: Burkina Faso, Côte

d’Ivoire, The Gambia, Ghana, Guinea, Niger, Nigeria, and Sierra Leone.14. Treaty Establishing the West African Economic and Monetary

Union, 1996, Articles 4, 91–93.15. Host countries typically act unilaterally in determining whether

temporary workers may bring their families and settle. Most Europeancountries allow temporary workers whose work permits have beenrenewed several times to obtain immigrant status after five years.

16. Between 1960 and 1973, the number of migrant workers inWestern Europe jumped from 2 million to 7 million, and the total for-eign population rose from 4 million to 12 million. Most of these work-ers came from geographically distant nations such as Turkey orMorocco, rather than neighboring countries. After the halt in temporaryworker programs in the mid-1970s, the migrant workforce in Europestabilized at around 5 million over the next decade. See Martin (2007);see also Council of Europe (1996).

17. The International Labour Organization (ILO) has developed amultilateral framework on labor migration that constitutes a compre-hensive collection of principles, guidelines, and good practices on labormigration programs, including bilateral labor agreements; see ILO,http://www.ilo.org/public/english/protection/migrant/areas/multilateral.htm.

18. A few memoranda of understanding on migratory and laborcooperation have been signed recently by developing countries; theyinclude those between Peru and Chile (2006), Peru and Ecuador(2006), and Peru and Mexico (2002). The aim of these memoranda is toprovide for exchange of information and protection of the rights of

migrant workers, in particular under the UN International Conventionon the Protection of the Rights of Migrant Workers and Their Relatives.The memoranda do not include provisions for promoting labor mobility.The Philippines has signed bilateral memoranda with many destinationcountries to cover the flows, rights, and obligations of its temporaryworkers. A reciprocal temporary worker program agreed on by Argentinaand Bolivia includes many of these protections.

19. As part of Canada’s SAWP scheme with Caribbean countries andMexico, the HRSDC cooperates closely with private agencies, includingForeign Agricultural Resource Management Services (FARMS) in Ontarioand Nova Scotia and the Foundation of Enterprises for the Recruitmentof Foreign Labor (FERME) in Quebec, New Brunswick, and PrinceEdward Island. Guest workers in Canada are employed in four provinces:Ontario (two-thirds of the total), Quebec, Alberta, and Manitoba.Although the Mexican government tries to ensure that every workerreturns to Mexico, independent researchers estimate that 15 percent of theMexicans fail to return home every year. See http://migration.ucdavis.edu/RMN/more.

20. Gao, personal communication (see note 9). The data are from theChina Foreign Labor Cooperation Annual Report 2004, issued by the ChinaInternational Contractors Association, Beijing.

21. See http://www.jitco.or.jp for details on this program. Informa-tion on China’s bilateral labor agreements was provided to the authors byDr. Shu Bin, manager, Labor Department, China National Aero-Technol-ogy Import & Export Corporation, during a workshop held in Beijingunder World Bank auspices, May 21–22, 2009.

22. A recent study by Persin (2008) compares the United Kingdom’sresponses concerning its labor market and immigration policies in thecontext of the eastern enlargement of the EU with its willingness to pro-vide offers on mode 4 under the GATS. Persin finds that the governmentopted for managed migration through bilateral labor agreements and anemployer-led system, rather than through more formal WTO commit-ments. The author concludes that the more flexible “bilateral or regionallabor immigration schemes are preferred to a binding multilateral laborimmigration scheme such as the GATS” because it is easier and less costlyunder the former to agree on rules and procedures, as well as to solve anyproblems jointly.

23. The Migration Policy Institute, based in Washington, DC, hasdone considerable work on the benefits and challenges of temporaryworker programs and circular migration schemes. See Batalova (2006);Meyers (2006); Agunias (2007); Newland, Agunias, and Terrazas (2008).

24. Unlike BLAs and BITs, DTTs are specifically exempted fromnational treatment and MFN requirements by GATS Article XIV.

Bibliography

Agunias, Dovelyn Rannveig. 2007. “Managing Temporary Migration: Lessons from the Philippine Model.” Insight. Migration Policy Institute,Washington, DC. http://www.migrationpolicy.org/pubs/Insight_POEA_Oct07.pdf.

Batalova, Jeanne. 2006. “The Growing Connection between Temporaryand Permanent Immigration Systems.” Task Force Insight 14. Migra-tion Policy Institute, Washington, DC. http://www.migrationpolicy.org/ITFIAF/TFI_Batalova.pdf.

Borjas, George J. 1999. Heaven’s Door: Immigration Policy and the Ameri-can Economy. Princeton, NJ: Princeton University Press.

Carzaniga, Antonia. 2008. “A Warmer Welcome? Access for Natural Persons under PTAs.” In Opening Markets for Trade in Services, ed.Juan A. Marchetti and Martin Roy, 475–502. London: CambridgeUniversity Press.

Cattaneo, Olivier, Michael Engman, Sebastián Sáez, and Robert M. Stern.2010. International Trade in Services: New Trends and Opportunities forDeveloping Countries. Washington, DC: World Bank.

Council of Europe. 1996. “Temporary Migration for Employment andTraining Purposes.” European Committee on Migration (CDMG),

Labor Mobility 305


Migration Policy Institute, Washington, DC. http://www.migrationpol-icy.org/pubs/Insight-IGC-Sept08.pdf.

Nielson, Julia. 2003. “Labor Mobility in Preferential Trade Agreements.” InMoving People to Deliver Services, ed. Aaditya Mattoo and AntoniaCarzaniga, 93–113. Washington, DC: World Bank.

OECD (Organisation for Economic Co-operation and Development).2007. International Migration Outlook. Paris: OECD.

———. 2008. West African Mobility and OECD Migration Policies.Paris: OECD. http://www.oecd.org/document/29/0,3343,en_38233741_38247095_41481445_1_1_1_1,00.html.

Persin, Daniela. 2008. “Free Movement of Labor: UK Responses to theEastern Enlargement and GATS Mode 4.” Journal of World Trade42 (5): 837–64.

Pritchett, Lant. 2006. Let Their People Come: Breaking the Gridlock on GlobalLabor Mobility. Washington, DC: Center for Global Development.

Sáez, Sebastian. 2009. “The EU EPA Negotiations.” Presented at a WorldBank training course on services, Washington, DC, March 21.

Soesastro, Hadi. 2005. “Accelerating ASEAN Integration: Moving BeyondAFTA.” CSIS Economics Working Paper Series 091, Centre for Strategicand International Studies, Jakarta. http://www.csis.or.id/publications_paper_view.asp?id=52&tab=1.

Stephenson, Sherry. 2007. “Mobility of Service Providers in the CaribbeanRegion.” Box 13.6 in L. Alan Winters, “The Temporary Movement ofWorkers to Provide Services: GATS (Mode 4).” In A Handbook of Inter-national Trade in Services, ed. Aaditya Mattoo, Robert M. Stern, andGianni Zanini, 535–37. Oxford, U.K.: Oxford University Press.

———. 2008. “Demographic Change and International Labor Mobility inthe Pacific Americas—Issues, Policies and Implications for Coopera-tion.” In Labor Mobility in the Asia-Pacific Region, ed. Graeme Hugo andSoogil Young, 171–201. Singapore: Institute of Southeast Asian Studies.

Vis-Dunbar, Damon, and Henrique Suzy Nikiema. 2009. “Do BilateralInvestment Treaties Lead to More Foreign Investment?” InvestmentTreaty News, April 30. International Institute for Sustainable Develop-ment, Winnipeg, Manitoba, Canada.

Walmsley, Terri Louise, and L. Alan Winters. 2002. “Relaxing the Restric-tions on the Temporary Movement of Natural Persons: A SimulationAnalysis.” CEPR Discussion Paper Series 3719, Centre for EconomicPolicy Research, London.

Winters, L. Alan. 2008. “The Temporary Movement of Workers to ProvideServices (GATS Mode 4).” In A Handbook of International Trade inServices, ed. Aaditya Mattoo, Robert M. Stern, and Gianni Zanini,480–96. Oxford, U.K.: Oxford University Press.

World Bank. 2006. Global Economic Prospects 2006: Economic Implicationsof Remittances and Migration. Washington, DC: World Bank.

———. 2010. Global Economic Prospects 2010: Crisis, Finance, andGrowth. Washington, DC: World Bank.

Strasbourg. http://www.coe.int/t/dg3/migration/Documentation/Legal_texts/CDMG_96_18e_Temporary_migration_employment_training_purposes_en.pdf.

Hamilton, Bob, and John Whalley. 1984. “Efficiency and DistributionalImplications of Global Restrictions on Labor Mobility.” Journal ofDevelopment Economics 14: 61–75.

Hatton, Timothy, and Jeffrey G. Williamson. 1998. The Age of Mass Migration: Causes and Economic Impact. New York and Oxford, U.K.:Oxford University Press.

IMF (International Monetary Fund). 2010. Balance of Payments Manual,Sixth Edition. Washington, DC: World Bank.

Iregui, Ana. 1999. “Efficiency Gains from the Elimination of GlobalRestrictions on Labor Mobility: An Analysis Using a MultiregionalCGE Model.” Estudios Economicos, Banco de la República, Bogota.

Karsenty, Guy. 2000. “Assessing Trade in Services by Mode of Supply.” InGATS 2000: New Directions in Services Trade Liberalization, ed. PierreSauvé and Robert M. Stern. Washington, DC: Brookings InstitutionPress.

Martin, Philip. 2007. Towards Effective Temporary Worker Programs: Issuesand Challenges in Industrial Countries. International Migration Papers89. Geneva: International Labour Office. http://www.ilo.org/public/english/protection/migrant/download/tempworkers_martin_en.pdf.

Martin, Susan, and B. Lindsay Lowell. 2008. “Examining Labor Mobil-ity Provisions for the United States.” Presented at the seminar on“Demographic Change and International Labor Mobility in theAsia Pacific Region: Implications for Business and Cooperation,”organized by Korea National Committee for Pacific Economic Coop-eration (KOPEC) and the Pacific Economic Cooperation Council(PECC), Seoul, March 24–26. http://www.pecc.org/component/eventlist/details/121-kopec-demographic-change-and-international-labor-mobility-in-the-asia-pacific-region.

Mattoo, Aaditya, and Antonia Carzaniga, eds. 2003. Moving People toDeliver Services. Washington, DC: World Bank.

Mattoo, Aaditya, and Pierre Sauvé. 2010. “Regionalism in Services Trade.”in A Handbook of International Trade in Services, ed. Aaditya Mattoo,Robert M. Stern, and Gianni Zanini. 2nd ed. Oxford, U.K.: OxfordUniversity Press.

Meyers, Deborah W. 2006. “Temporary Worker Programs: A PatchworkPolicy Response.” Task Force Insight 12, Migration Policy Institute,Washington, DC. http://www.migrationpolicy.org/ITFIAF/TFI_12_Meyers.pdf.

Moses, Jonathon, and Bjorn Letnes. 2004. “The Economic Cost to Inter-national Labor Restrictions: Revisiting the Empirical Discussion.”World Development 32 (10): 1609–26.

Newland, Kathleen, Dovelyn Rannveig Agunias, and Aaron Terrazas. 2008. “Learning by Doing: Experiences of Circular Migration.” Insight,



Documents

inte 001-028 ch01 - World Banksiteresources.worldbank.org/INTRANETTRADE/Resources/C13.pdf · ubp_PTAPD_275-306.qxd:inte_001-028_ch01 7/5/11 12:28 PM Page 275. The Concept of Labor